checkpoint step 400
Browse files- .gitattributes +1 -0
- checkpoint-400/README.md +209 -0
- checkpoint-400/adapter_config.json +48 -0
- checkpoint-400/adapter_model.safetensors +3 -0
- checkpoint-400/chat_template.jinja +54 -0
- checkpoint-400/optimizer.pt +3 -0
- checkpoint-400/rng_state.pth +3 -0
- checkpoint-400/scheduler.pt +3 -0
- checkpoint-400/tokenizer.json +3 -0
- checkpoint-400/tokenizer_config.json +16 -0
- checkpoint-400/trainer_state.json +2194 -0
- checkpoint-400/training_args.bin +3 -0
.gitattributes
CHANGED
|
@@ -40,3 +40,4 @@ checkpoint-200/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
|
| 40 |
checkpoint-250/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 41 |
checkpoint-300/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 42 |
checkpoint-350/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 40 |
checkpoint-250/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 41 |
checkpoint-300/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 42 |
checkpoint-350/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 43 |
+
checkpoint-400/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
checkpoint-400/README.md
ADDED
|
@@ -0,0 +1,209 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: unsloth/Qwen2.5-7B-Instruct
|
| 3 |
+
library_name: peft
|
| 4 |
+
pipeline_tag: text-generation
|
| 5 |
+
tags:
|
| 6 |
+
- base_model:adapter:unsloth/Qwen2.5-7B-Instruct
|
| 7 |
+
- grpo
|
| 8 |
+
- lora
|
| 9 |
+
- transformers
|
| 10 |
+
- trl
|
| 11 |
+
---
|
| 12 |
+
|
| 13 |
+
# Model Card for Model ID
|
| 14 |
+
|
| 15 |
+
<!-- Provide a quick summary of what the model is/does. -->
|
| 16 |
+
|
| 17 |
+
|
| 18 |
+
|
| 19 |
+
## Model Details
|
| 20 |
+
|
| 21 |
+
### Model Description
|
| 22 |
+
|
| 23 |
+
<!-- Provide a longer summary of what this model is. -->
|
| 24 |
+
|
| 25 |
+
|
| 26 |
+
|
| 27 |
+
- **Developed by:** [More Information Needed]
|
| 28 |
+
- **Funded by [optional]:** [More Information Needed]
|
| 29 |
+
- **Shared by [optional]:** [More Information Needed]
|
| 30 |
+
- **Model type:** [More Information Needed]
|
| 31 |
+
- **Language(s) (NLP):** [More Information Needed]
|
| 32 |
+
- **License:** [More Information Needed]
|
| 33 |
+
- **Finetuned from model [optional]:** [More Information Needed]
|
| 34 |
+
|
| 35 |
+
### Model Sources [optional]
|
| 36 |
+
|
| 37 |
+
<!-- Provide the basic links for the model. -->
|
| 38 |
+
|
| 39 |
+
- **Repository:** [More Information Needed]
|
| 40 |
+
- **Paper [optional]:** [More Information Needed]
|
| 41 |
+
- **Demo [optional]:** [More Information Needed]
|
| 42 |
+
|
| 43 |
+
## Uses
|
| 44 |
+
|
| 45 |
+
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
|
| 46 |
+
|
| 47 |
+
### Direct Use
|
| 48 |
+
|
| 49 |
+
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
|
| 50 |
+
|
| 51 |
+
[More Information Needed]
|
| 52 |
+
|
| 53 |
+
### Downstream Use [optional]
|
| 54 |
+
|
| 55 |
+
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
|
| 56 |
+
|
| 57 |
+
[More Information Needed]
|
| 58 |
+
|
| 59 |
+
### Out-of-Scope Use
|
| 60 |
+
|
| 61 |
+
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
|
| 62 |
+
|
| 63 |
+
[More Information Needed]
|
| 64 |
+
|
| 65 |
+
## Bias, Risks, and Limitations
|
| 66 |
+
|
| 67 |
+
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
|
| 68 |
+
|
| 69 |
+
[More Information Needed]
|
| 70 |
+
|
| 71 |
+
### Recommendations
|
| 72 |
+
|
| 73 |
+
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
|
| 74 |
+
|
| 75 |
+
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
|
| 76 |
+
|
| 77 |
+
## How to Get Started with the Model
|
| 78 |
+
|
| 79 |
+
Use the code below to get started with the model.
|
| 80 |
+
|
| 81 |
+
[More Information Needed]
|
| 82 |
+
|
| 83 |
+
## Training Details
|
| 84 |
+
|
| 85 |
+
### Training Data
|
| 86 |
+
|
| 87 |
+
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
|
| 88 |
+
|
| 89 |
+
[More Information Needed]
|
| 90 |
+
|
| 91 |
+
### Training Procedure
|
| 92 |
+
|
| 93 |
+
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
|
| 94 |
+
|
| 95 |
+
#### Preprocessing [optional]
|
| 96 |
+
|
| 97 |
+
[More Information Needed]
|
| 98 |
+
|
| 99 |
+
|
| 100 |
+
#### Training Hyperparameters
|
| 101 |
+
|
| 102 |
+
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
|
| 103 |
+
|
| 104 |
+
#### Speeds, Sizes, Times [optional]
|
| 105 |
+
|
| 106 |
+
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
|
| 107 |
+
|
| 108 |
+
[More Information Needed]
|
| 109 |
+
|
| 110 |
+
## Evaluation
|
| 111 |
+
|
| 112 |
+
<!-- This section describes the evaluation protocols and provides the results. -->
|
| 113 |
+
|
| 114 |
+
### Testing Data, Factors & Metrics
|
| 115 |
+
|
| 116 |
+
#### Testing Data
|
| 117 |
+
|
| 118 |
+
<!-- This should link to a Dataset Card if possible. -->
|
| 119 |
+
|
| 120 |
+
[More Information Needed]
|
| 121 |
+
|
| 122 |
+
#### Factors
|
| 123 |
+
|
| 124 |
+
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
|
| 125 |
+
|
| 126 |
+
[More Information Needed]
|
| 127 |
+
|
| 128 |
+
#### Metrics
|
| 129 |
+
|
| 130 |
+
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
| 131 |
+
|
| 132 |
+
[More Information Needed]
|
| 133 |
+
|
| 134 |
+
### Results
|
| 135 |
+
|
| 136 |
+
[More Information Needed]
|
| 137 |
+
|
| 138 |
+
#### Summary
|
| 139 |
+
|
| 140 |
+
|
| 141 |
+
|
| 142 |
+
## Model Examination [optional]
|
| 143 |
+
|
| 144 |
+
<!-- Relevant interpretability work for the model goes here -->
|
| 145 |
+
|
| 146 |
+
[More Information Needed]
|
| 147 |
+
|
| 148 |
+
## Environmental Impact
|
| 149 |
+
|
| 150 |
+
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
|
| 151 |
+
|
| 152 |
+
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
|
| 153 |
+
|
| 154 |
+
- **Hardware Type:** [More Information Needed]
|
| 155 |
+
- **Hours used:** [More Information Needed]
|
| 156 |
+
- **Cloud Provider:** [More Information Needed]
|
| 157 |
+
- **Compute Region:** [More Information Needed]
|
| 158 |
+
- **Carbon Emitted:** [More Information Needed]
|
| 159 |
+
|
| 160 |
+
## Technical Specifications [optional]
|
| 161 |
+
|
| 162 |
+
### Model Architecture and Objective
|
| 163 |
+
|
| 164 |
+
[More Information Needed]
|
| 165 |
+
|
| 166 |
+
### Compute Infrastructure
|
| 167 |
+
|
| 168 |
+
[More Information Needed]
|
| 169 |
+
|
| 170 |
+
#### Hardware
|
| 171 |
+
|
| 172 |
+
[More Information Needed]
|
| 173 |
+
|
| 174 |
+
#### Software
|
| 175 |
+
|
| 176 |
+
[More Information Needed]
|
| 177 |
+
|
| 178 |
+
## Citation [optional]
|
| 179 |
+
|
| 180 |
+
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
|
| 181 |
+
|
| 182 |
+
**BibTeX:**
|
| 183 |
+
|
| 184 |
+
[More Information Needed]
|
| 185 |
+
|
| 186 |
+
**APA:**
|
| 187 |
+
|
| 188 |
+
[More Information Needed]
|
| 189 |
+
|
| 190 |
+
## Glossary [optional]
|
| 191 |
+
|
| 192 |
+
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
|
| 193 |
+
|
| 194 |
+
[More Information Needed]
|
| 195 |
+
|
| 196 |
+
## More Information [optional]
|
| 197 |
+
|
| 198 |
+
[More Information Needed]
|
| 199 |
+
|
| 200 |
+
## Model Card Authors [optional]
|
| 201 |
+
|
| 202 |
+
[More Information Needed]
|
| 203 |
+
|
| 204 |
+
## Model Card Contact
|
| 205 |
+
|
| 206 |
+
[More Information Needed]
|
| 207 |
+
### Framework versions
|
| 208 |
+
|
| 209 |
+
- PEFT 0.19.1
|
checkpoint-400/adapter_config.json
ADDED
|
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alora_invocation_tokens": null,
|
| 3 |
+
"alpha_pattern": {},
|
| 4 |
+
"arrow_config": null,
|
| 5 |
+
"auto_mapping": null,
|
| 6 |
+
"base_model_name_or_path": "unsloth/Qwen2.5-7B-Instruct",
|
| 7 |
+
"bias": "none",
|
| 8 |
+
"corda_config": null,
|
| 9 |
+
"ensure_weight_tying": false,
|
| 10 |
+
"eva_config": null,
|
| 11 |
+
"exclude_modules": null,
|
| 12 |
+
"fan_in_fan_out": false,
|
| 13 |
+
"inference_mode": true,
|
| 14 |
+
"init_lora_weights": true,
|
| 15 |
+
"layer_replication": null,
|
| 16 |
+
"layers_pattern": null,
|
| 17 |
+
"layers_to_transform": null,
|
| 18 |
+
"loftq_config": {},
|
| 19 |
+
"lora_alpha": 16,
|
| 20 |
+
"lora_bias": false,
|
| 21 |
+
"lora_dropout": 0.05,
|
| 22 |
+
"lora_ga_config": null,
|
| 23 |
+
"megatron_config": null,
|
| 24 |
+
"megatron_core": "megatron.core",
|
| 25 |
+
"modules_to_save": null,
|
| 26 |
+
"peft_type": "LORA",
|
| 27 |
+
"peft_version": "0.19.1",
|
| 28 |
+
"qalora_group_size": 16,
|
| 29 |
+
"r": 16,
|
| 30 |
+
"rank_pattern": {},
|
| 31 |
+
"revision": null,
|
| 32 |
+
"target_modules": [
|
| 33 |
+
"down_proj",
|
| 34 |
+
"up_proj",
|
| 35 |
+
"k_proj",
|
| 36 |
+
"v_proj",
|
| 37 |
+
"gate_proj",
|
| 38 |
+
"q_proj",
|
| 39 |
+
"o_proj"
|
| 40 |
+
],
|
| 41 |
+
"target_parameters": null,
|
| 42 |
+
"task_type": "CAUSAL_LM",
|
| 43 |
+
"trainable_token_indices": null,
|
| 44 |
+
"use_bdlora": null,
|
| 45 |
+
"use_dora": false,
|
| 46 |
+
"use_qalora": false,
|
| 47 |
+
"use_rslora": false
|
| 48 |
+
}
|
checkpoint-400/adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:239cdb5da5d66fdb33a4d7a26c1cc2849e797d232a44da1637b9fdf6ec4128f1
|
| 3 |
+
size 80792880
|
checkpoint-400/chat_template.jinja
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{%- if tools %}
|
| 2 |
+
{{- '<|im_start|>system\n' }}
|
| 3 |
+
{%- if messages[0]['role'] == 'system' %}
|
| 4 |
+
{{- messages[0]['content'] }}
|
| 5 |
+
{%- else %}
|
| 6 |
+
{{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
|
| 7 |
+
{%- endif %}
|
| 8 |
+
{{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
|
| 9 |
+
{%- for tool in tools %}
|
| 10 |
+
{{- "\n" }}
|
| 11 |
+
{{- tool | tojson }}
|
| 12 |
+
{%- endfor %}
|
| 13 |
+
{{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
|
| 14 |
+
{%- else %}
|
| 15 |
+
{%- if messages[0]['role'] == 'system' %}
|
| 16 |
+
{{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
|
| 17 |
+
{%- else %}
|
| 18 |
+
{{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
|
| 19 |
+
{%- endif %}
|
| 20 |
+
{%- endif %}
|
| 21 |
+
{%- for message in messages %}
|
| 22 |
+
{%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
|
| 23 |
+
{{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
|
| 24 |
+
{%- elif message.role == "assistant" %}
|
| 25 |
+
{{- '<|im_start|>' + message.role }}
|
| 26 |
+
{%- if message.content %}
|
| 27 |
+
{{- '\n' + message.content }}
|
| 28 |
+
{%- endif %}
|
| 29 |
+
{%- for tool_call in message.tool_calls %}
|
| 30 |
+
{%- if tool_call.function is defined %}
|
| 31 |
+
{%- set tool_call = tool_call.function %}
|
| 32 |
+
{%- endif %}
|
| 33 |
+
{{- '\n<tool_call>\n{"name": "' }}
|
| 34 |
+
{{- tool_call.name }}
|
| 35 |
+
{{- '", "arguments": ' }}
|
| 36 |
+
{{- tool_call.arguments | tojson }}
|
| 37 |
+
{{- '}\n</tool_call>' }}
|
| 38 |
+
{%- endfor %}
|
| 39 |
+
{{- '<|im_end|>\n' }}
|
| 40 |
+
{%- elif message.role == "tool" %}
|
| 41 |
+
{%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
|
| 42 |
+
{{- '<|im_start|>user' }}
|
| 43 |
+
{%- endif %}
|
| 44 |
+
{{- '\n<tool_response>\n' }}
|
| 45 |
+
{{- message.content }}
|
| 46 |
+
{{- '\n</tool_response>' }}
|
| 47 |
+
{%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
|
| 48 |
+
{{- '<|im_end|>\n' }}
|
| 49 |
+
{%- endif %}
|
| 50 |
+
{%- endif %}
|
| 51 |
+
{%- endfor %}
|
| 52 |
+
{%- if add_generation_prompt %}
|
| 53 |
+
{{- '<|im_start|>assistant\n' }}
|
| 54 |
+
{%- endif %}
|
checkpoint-400/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bcfb16bafb5eb571814c955f24eeae1f8a3497a5ac8705ac5c54000d8fc57db7
|
| 3 |
+
size 161816251
|
checkpoint-400/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ef1631048dda7f15e4e6cbb07af1bb611876333139b2bdee6c0ac6a43f75a273
|
| 3 |
+
size 14645
|
checkpoint-400/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:da90f75114b53b992be47499086dc4b49081c2957d7aa7214da1e6d3efb54ba3
|
| 3 |
+
size 1465
|
checkpoint-400/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
|
| 3 |
+
size 11421892
|
checkpoint-400/tokenizer_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": null,
|
| 5 |
+
"clean_up_tokenization_spaces": false,
|
| 6 |
+
"eos_token": "<|im_end|>",
|
| 7 |
+
"errors": "replace",
|
| 8 |
+
"is_local": false,
|
| 9 |
+
"local_files_only": false,
|
| 10 |
+
"model_max_length": 32768,
|
| 11 |
+
"pad_token": "<|vision_pad|>",
|
| 12 |
+
"padding_side": "left",
|
| 13 |
+
"split_special_tokens": false,
|
| 14 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 15 |
+
"unk_token": null
|
| 16 |
+
}
|
checkpoint-400/trainer_state.json
ADDED
|
@@ -0,0 +1,2194 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.017241379310344827,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 400,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"clip_ratio/high_max": 0.0,
|
| 14 |
+
"clip_ratio/high_mean": 0.0,
|
| 15 |
+
"clip_ratio/low_mean": 0.0,
|
| 16 |
+
"clip_ratio/low_min": 0.0,
|
| 17 |
+
"clip_ratio/region_mean": 0.0,
|
| 18 |
+
"completions/clipped_ratio": 0.0,
|
| 19 |
+
"completions/max_length": 130.0,
|
| 20 |
+
"completions/max_terminated_length": 130.0,
|
| 21 |
+
"completions/mean_length": 68.5,
|
| 22 |
+
"completions/mean_terminated_length": 68.5,
|
| 23 |
+
"completions/min_length": 26.8,
|
| 24 |
+
"completions/min_terminated_length": 26.8,
|
| 25 |
+
"entropy": 0.25826080311089755,
|
| 26 |
+
"epoch": 0.00021551724137931034,
|
| 27 |
+
"frac_reward_zero_std": 0.1,
|
| 28 |
+
"grad_norm": 0.28125,
|
| 29 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 30 |
+
"loss": 0.10238287448883057,
|
| 31 |
+
"num_tokens": 12064.0,
|
| 32 |
+
"reward": 0.3290319949388504,
|
| 33 |
+
"reward_std": 0.40028320252895355,
|
| 34 |
+
"rewards/reward_fn/mean": 0.3290319949388504,
|
| 35 |
+
"rewards/reward_fn/std": 0.400283208489418,
|
| 36 |
+
"step": 5,
|
| 37 |
+
"step_time": 22.194597821800016
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"clip_ratio/high_max": 0.0,
|
| 41 |
+
"clip_ratio/high_mean": 0.0,
|
| 42 |
+
"clip_ratio/low_mean": 0.0,
|
| 43 |
+
"clip_ratio/low_min": 0.0,
|
| 44 |
+
"clip_ratio/region_mean": 0.0,
|
| 45 |
+
"completions/clipped_ratio": 0.0,
|
| 46 |
+
"completions/max_length": 87.6,
|
| 47 |
+
"completions/max_terminated_length": 87.6,
|
| 48 |
+
"completions/mean_length": 48.525,
|
| 49 |
+
"completions/mean_terminated_length": 48.525,
|
| 50 |
+
"completions/min_length": 17.0,
|
| 51 |
+
"completions/min_terminated_length": 17.0,
|
| 52 |
+
"entropy": 0.2673827801831067,
|
| 53 |
+
"epoch": 0.0004310344827586207,
|
| 54 |
+
"frac_reward_zero_std": 0.1,
|
| 55 |
+
"grad_norm": 0.484375,
|
| 56 |
+
"learning_rate": 2.25e-06,
|
| 57 |
+
"loss": 0.05849265456199646,
|
| 58 |
+
"num_tokens": 23089.0,
|
| 59 |
+
"reward": 0.45969198942184447,
|
| 60 |
+
"reward_std": 0.2795014828443527,
|
| 61 |
+
"rewards/reward_fn/mean": 0.45969198942184447,
|
| 62 |
+
"rewards/reward_fn/std": 0.27950150668621065,
|
| 63 |
+
"step": 10,
|
| 64 |
+
"step_time": 16.331819552399928
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"clip_ratio/high_max": 0.0,
|
| 68 |
+
"clip_ratio/high_mean": 0.0,
|
| 69 |
+
"clip_ratio/low_mean": 0.0,
|
| 70 |
+
"clip_ratio/low_min": 0.0,
|
| 71 |
+
"clip_ratio/region_mean": 0.0,
|
| 72 |
+
"completions/clipped_ratio": 0.0,
|
| 73 |
+
"completions/max_length": 85.6,
|
| 74 |
+
"completions/max_terminated_length": 85.6,
|
| 75 |
+
"completions/mean_length": 47.35,
|
| 76 |
+
"completions/mean_terminated_length": 47.35,
|
| 77 |
+
"completions/min_length": 25.0,
|
| 78 |
+
"completions/min_terminated_length": 25.0,
|
| 79 |
+
"entropy": 0.20174795808270574,
|
| 80 |
+
"epoch": 0.000646551724137931,
|
| 81 |
+
"frac_reward_zero_std": 0.4,
|
| 82 |
+
"grad_norm": 0.66015625,
|
| 83 |
+
"learning_rate": 3.5e-06,
|
| 84 |
+
"loss": -0.03555725216865539,
|
| 85 |
+
"num_tokens": 34239.0,
|
| 86 |
+
"reward": 0.5545999944210053,
|
| 87 |
+
"reward_std": 0.32832055240869523,
|
| 88 |
+
"rewards/reward_fn/mean": 0.5545999944210053,
|
| 89 |
+
"rewards/reward_fn/std": 0.3283205583691597,
|
| 90 |
+
"step": 15,
|
| 91 |
+
"step_time": 16.44816417159991
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"clip_ratio/high_max": 0.0,
|
| 95 |
+
"clip_ratio/high_mean": 0.0,
|
| 96 |
+
"clip_ratio/low_mean": 0.0,
|
| 97 |
+
"clip_ratio/low_min": 0.0,
|
| 98 |
+
"clip_ratio/region_mean": 0.0,
|
| 99 |
+
"completions/clipped_ratio": 0.0,
|
| 100 |
+
"completions/max_length": 98.2,
|
| 101 |
+
"completions/max_terminated_length": 98.2,
|
| 102 |
+
"completions/mean_length": 53.6,
|
| 103 |
+
"completions/mean_terminated_length": 53.6,
|
| 104 |
+
"completions/min_length": 25.0,
|
| 105 |
+
"completions/min_terminated_length": 25.0,
|
| 106 |
+
"entropy": 0.2620095924474299,
|
| 107 |
+
"epoch": 0.0008620689655172414,
|
| 108 |
+
"frac_reward_zero_std": 0.2,
|
| 109 |
+
"grad_norm": 0.5390625,
|
| 110 |
+
"learning_rate": 4.75e-06,
|
| 111 |
+
"loss": 0.017401468753814698,
|
| 112 |
+
"num_tokens": 45399.0,
|
| 113 |
+
"reward": 0.4466240078210831,
|
| 114 |
+
"reward_std": 0.27573536019772293,
|
| 115 |
+
"rewards/reward_fn/mean": 0.4466240078210831,
|
| 116 |
+
"rewards/reward_fn/std": 0.27573537137359383,
|
| 117 |
+
"step": 20,
|
| 118 |
+
"step_time": 17.988385831799953
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"clip_ratio/high_max": 0.0,
|
| 122 |
+
"clip_ratio/high_mean": 0.0,
|
| 123 |
+
"clip_ratio/low_mean": 0.0,
|
| 124 |
+
"clip_ratio/low_min": 0.0,
|
| 125 |
+
"clip_ratio/region_mean": 0.0,
|
| 126 |
+
"completions/clipped_ratio": 0.0,
|
| 127 |
+
"completions/max_length": 123.8,
|
| 128 |
+
"completions/max_terminated_length": 123.8,
|
| 129 |
+
"completions/mean_length": 59.15,
|
| 130 |
+
"completions/mean_terminated_length": 59.15,
|
| 131 |
+
"completions/min_length": 21.8,
|
| 132 |
+
"completions/min_terminated_length": 21.8,
|
| 133 |
+
"entropy": 0.33701689867302775,
|
| 134 |
+
"epoch": 0.0010775862068965517,
|
| 135 |
+
"frac_reward_zero_std": 0.0,
|
| 136 |
+
"grad_norm": 0.74609375,
|
| 137 |
+
"learning_rate": 4.965517241379311e-06,
|
| 138 |
+
"loss": -0.02674364447593689,
|
| 139 |
+
"num_tokens": 56993.0,
|
| 140 |
+
"reward": 0.5006439983844757,
|
| 141 |
+
"reward_std": 0.2351181447505951,
|
| 142 |
+
"rewards/reward_fn/mean": 0.5006439983844757,
|
| 143 |
+
"rewards/reward_fn/std": 0.23511814773082734,
|
| 144 |
+
"step": 25,
|
| 145 |
+
"step_time": 21.11461545459997
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"clip_ratio/high_max": 0.0,
|
| 149 |
+
"clip_ratio/high_mean": 0.0,
|
| 150 |
+
"clip_ratio/low_mean": 0.0,
|
| 151 |
+
"clip_ratio/low_min": 0.0,
|
| 152 |
+
"clip_ratio/region_mean": 0.0,
|
| 153 |
+
"completions/clipped_ratio": 0.0,
|
| 154 |
+
"completions/max_length": 110.2,
|
| 155 |
+
"completions/max_terminated_length": 110.2,
|
| 156 |
+
"completions/mean_length": 47.95,
|
| 157 |
+
"completions/mean_terminated_length": 47.95,
|
| 158 |
+
"completions/min_length": 15.8,
|
| 159 |
+
"completions/min_terminated_length": 15.8,
|
| 160 |
+
"entropy": 0.24504410615190864,
|
| 161 |
+
"epoch": 0.001293103448275862,
|
| 162 |
+
"frac_reward_zero_std": 0.4,
|
| 163 |
+
"grad_norm": 0.92578125,
|
| 164 |
+
"learning_rate": 4.922413793103449e-06,
|
| 165 |
+
"loss": -0.05023183822631836,
|
| 166 |
+
"num_tokens": 67659.0,
|
| 167 |
+
"reward": 0.6457239985466003,
|
| 168 |
+
"reward_std": 0.11179900387069211,
|
| 169 |
+
"rewards/reward_fn/mean": 0.6457239985466003,
|
| 170 |
+
"rewards/reward_fn/std": 0.11179901438299567,
|
| 171 |
+
"step": 30,
|
| 172 |
+
"step_time": 19.134268762199827
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"clip_ratio/high_max": 0.0,
|
| 176 |
+
"clip_ratio/high_mean": 0.0,
|
| 177 |
+
"clip_ratio/low_mean": 0.0,
|
| 178 |
+
"clip_ratio/low_min": 0.0,
|
| 179 |
+
"clip_ratio/region_mean": 0.0,
|
| 180 |
+
"completions/clipped_ratio": 0.1,
|
| 181 |
+
"completions/max_length": 150.8,
|
| 182 |
+
"completions/max_terminated_length": 120.2,
|
| 183 |
+
"completions/mean_length": 73.375,
|
| 184 |
+
"completions/mean_terminated_length": 54.36666717529297,
|
| 185 |
+
"completions/min_length": 15.2,
|
| 186 |
+
"completions/min_terminated_length": 15.2,
|
| 187 |
+
"entropy": 0.23285924410447478,
|
| 188 |
+
"epoch": 0.0015086206896551724,
|
| 189 |
+
"frac_reward_zero_std": 0.3,
|
| 190 |
+
"grad_norm": 0.58203125,
|
| 191 |
+
"learning_rate": 4.879310344827586e-06,
|
| 192 |
+
"loss": -0.05149807929992676,
|
| 193 |
+
"num_tokens": 79734.0,
|
| 194 |
+
"reward": 0.46414998471736907,
|
| 195 |
+
"reward_std": 0.4071369742392562,
|
| 196 |
+
"rewards/reward_fn/mean": 0.46414998471736907,
|
| 197 |
+
"rewards/reward_fn/std": 0.40713699695188554,
|
| 198 |
+
"step": 35,
|
| 199 |
+
"step_time": 24.679699858600042
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"clip_ratio/high_max": 0.0,
|
| 203 |
+
"clip_ratio/high_mean": 0.0,
|
| 204 |
+
"clip_ratio/low_mean": 0.0,
|
| 205 |
+
"clip_ratio/low_min": 0.0,
|
| 206 |
+
"clip_ratio/region_mean": 0.0,
|
| 207 |
+
"completions/clipped_ratio": 0.0,
|
| 208 |
+
"completions/max_length": 125.4,
|
| 209 |
+
"completions/max_terminated_length": 125.4,
|
| 210 |
+
"completions/mean_length": 68.925,
|
| 211 |
+
"completions/mean_terminated_length": 68.925,
|
| 212 |
+
"completions/min_length": 26.0,
|
| 213 |
+
"completions/min_terminated_length": 26.0,
|
| 214 |
+
"entropy": 0.3110098702833056,
|
| 215 |
+
"epoch": 0.0017241379310344827,
|
| 216 |
+
"frac_reward_zero_std": 0.1,
|
| 217 |
+
"grad_norm": 0.419921875,
|
| 218 |
+
"learning_rate": 4.836206896551724e-06,
|
| 219 |
+
"loss": 0.030162644386291505,
|
| 220 |
+
"num_tokens": 91695.0,
|
| 221 |
+
"reward": 0.5143539935350419,
|
| 222 |
+
"reward_std": 0.2790891878306866,
|
| 223 |
+
"rewards/reward_fn/mean": 0.5143539935350419,
|
| 224 |
+
"rewards/reward_fn/std": 0.279089218378067,
|
| 225 |
+
"step": 40,
|
| 226 |
+
"step_time": 21.565513409599998
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"clip_ratio/high_max": 0.0,
|
| 230 |
+
"clip_ratio/high_mean": 0.0,
|
| 231 |
+
"clip_ratio/low_mean": 0.0,
|
| 232 |
+
"clip_ratio/low_min": 0.0,
|
| 233 |
+
"clip_ratio/region_mean": 0.0,
|
| 234 |
+
"completions/clipped_ratio": 0.0,
|
| 235 |
+
"completions/max_length": 100.0,
|
| 236 |
+
"completions/max_terminated_length": 100.0,
|
| 237 |
+
"completions/mean_length": 53.875,
|
| 238 |
+
"completions/mean_terminated_length": 53.875,
|
| 239 |
+
"completions/min_length": 16.6,
|
| 240 |
+
"completions/min_terminated_length": 16.6,
|
| 241 |
+
"entropy": 0.23047098610550165,
|
| 242 |
+
"epoch": 0.001939655172413793,
|
| 243 |
+
"frac_reward_zero_std": 0.4,
|
| 244 |
+
"grad_norm": 1.15625,
|
| 245 |
+
"learning_rate": 4.793103448275862e-06,
|
| 246 |
+
"loss": -0.04350074529647827,
|
| 247 |
+
"num_tokens": 102994.0,
|
| 248 |
+
"reward": 0.6398920059204102,
|
| 249 |
+
"reward_std": 0.2279826147481799,
|
| 250 |
+
"rewards/reward_fn/mean": 0.6398920059204102,
|
| 251 |
+
"rewards/reward_fn/std": 0.22798261437565087,
|
| 252 |
+
"step": 45,
|
| 253 |
+
"step_time": 18.23975218899968
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"clip_ratio/high_max": 0.0,
|
| 257 |
+
"clip_ratio/high_mean": 0.0,
|
| 258 |
+
"clip_ratio/low_mean": 0.0,
|
| 259 |
+
"clip_ratio/low_min": 0.0,
|
| 260 |
+
"clip_ratio/region_mean": 0.0,
|
| 261 |
+
"completions/clipped_ratio": 0.0,
|
| 262 |
+
"completions/max_length": 133.4,
|
| 263 |
+
"completions/max_terminated_length": 133.4,
|
| 264 |
+
"completions/mean_length": 75.975,
|
| 265 |
+
"completions/mean_terminated_length": 75.975,
|
| 266 |
+
"completions/min_length": 23.8,
|
| 267 |
+
"completions/min_terminated_length": 23.8,
|
| 268 |
+
"entropy": 0.34613882582634686,
|
| 269 |
+
"epoch": 0.0021551724137931034,
|
| 270 |
+
"frac_reward_zero_std": 0.2,
|
| 271 |
+
"grad_norm": 0.1318359375,
|
| 272 |
+
"learning_rate": 4.75e-06,
|
| 273 |
+
"loss": -0.03668657541275024,
|
| 274 |
+
"num_tokens": 115289.0,
|
| 275 |
+
"reward": 0.33042599707841874,
|
| 276 |
+
"reward_std": 0.5918701648712158,
|
| 277 |
+
"rewards/reward_fn/mean": 0.33042599707841874,
|
| 278 |
+
"rewards/reward_fn/std": 0.5918702006340026,
|
| 279 |
+
"step": 50,
|
| 280 |
+
"step_time": 22.43372436519985
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"clip_ratio/high_max": 0.0,
|
| 284 |
+
"clip_ratio/high_mean": 0.0,
|
| 285 |
+
"clip_ratio/low_mean": 0.0,
|
| 286 |
+
"clip_ratio/low_min": 0.0,
|
| 287 |
+
"clip_ratio/region_mean": 0.0,
|
| 288 |
+
"completions/clipped_ratio": 0.0,
|
| 289 |
+
"completions/max_length": 92.2,
|
| 290 |
+
"completions/max_terminated_length": 92.2,
|
| 291 |
+
"completions/mean_length": 50.65,
|
| 292 |
+
"completions/mean_terminated_length": 50.65,
|
| 293 |
+
"completions/min_length": 18.0,
|
| 294 |
+
"completions/min_terminated_length": 18.0,
|
| 295 |
+
"entropy": 0.3391244841972366,
|
| 296 |
+
"epoch": 0.0023706896551724138,
|
| 297 |
+
"frac_reward_zero_std": 0.3,
|
| 298 |
+
"grad_norm": 0.39453125,
|
| 299 |
+
"learning_rate": 4.706896551724138e-06,
|
| 300 |
+
"loss": -0.0521969735622406,
|
| 301 |
+
"num_tokens": 126111.0,
|
| 302 |
+
"reward": 0.6139639973640442,
|
| 303 |
+
"reward_std": 0.18208687230944634,
|
| 304 |
+
"rewards/reward_fn/mean": 0.6139639973640442,
|
| 305 |
+
"rewards/reward_fn/std": 0.18208688013255597,
|
| 306 |
+
"step": 55,
|
| 307 |
+
"step_time": 16.889007100199933
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"clip_ratio/high_max": 0.0,
|
| 311 |
+
"clip_ratio/high_mean": 0.0,
|
| 312 |
+
"clip_ratio/low_mean": 0.0,
|
| 313 |
+
"clip_ratio/low_min": 0.0,
|
| 314 |
+
"clip_ratio/region_mean": 0.0,
|
| 315 |
+
"completions/clipped_ratio": 0.025,
|
| 316 |
+
"completions/max_length": 164.0,
|
| 317 |
+
"completions/max_terminated_length": 162.4,
|
| 318 |
+
"completions/mean_length": 82.1,
|
| 319 |
+
"completions/mean_terminated_length": 78.87857208251953,
|
| 320 |
+
"completions/min_length": 29.4,
|
| 321 |
+
"completions/min_terminated_length": 29.4,
|
| 322 |
+
"entropy": 0.28240502553526314,
|
| 323 |
+
"epoch": 0.002586206896551724,
|
| 324 |
+
"frac_reward_zero_std": 0.5,
|
| 325 |
+
"grad_norm": 0.2373046875,
|
| 326 |
+
"learning_rate": 4.663793103448276e-06,
|
| 327 |
+
"loss": -0.0108717679977417,
|
| 328 |
+
"num_tokens": 138667.0,
|
| 329 |
+
"reward": 0.6616780042648316,
|
| 330 |
+
"reward_std": 0.18088504523038865,
|
| 331 |
+
"rewards/reward_fn/mean": 0.6616780042648316,
|
| 332 |
+
"rewards/reward_fn/std": 0.18088504672050476,
|
| 333 |
+
"step": 60,
|
| 334 |
+
"step_time": 26.46498533039994
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"clip_ratio/high_max": 0.0,
|
| 338 |
+
"clip_ratio/high_mean": 0.0,
|
| 339 |
+
"clip_ratio/low_mean": 0.0,
|
| 340 |
+
"clip_ratio/low_min": 0.0,
|
| 341 |
+
"clip_ratio/region_mean": 0.0,
|
| 342 |
+
"completions/clipped_ratio": 0.05,
|
| 343 |
+
"completions/max_length": 134.8,
|
| 344 |
+
"completions/max_terminated_length": 133.0,
|
| 345 |
+
"completions/mean_length": 69.55,
|
| 346 |
+
"completions/mean_terminated_length": 63.233334350585935,
|
| 347 |
+
"completions/min_length": 15.8,
|
| 348 |
+
"completions/min_terminated_length": 15.8,
|
| 349 |
+
"entropy": 0.3688268234953284,
|
| 350 |
+
"epoch": 0.0028017241379310344,
|
| 351 |
+
"frac_reward_zero_std": 0.1,
|
| 352 |
+
"grad_norm": 0.2216796875,
|
| 353 |
+
"learning_rate": 4.620689655172414e-06,
|
| 354 |
+
"loss": 0.09875075221061706,
|
| 355 |
+
"num_tokens": 150585.0,
|
| 356 |
+
"reward": 0.5008320093154908,
|
| 357 |
+
"reward_std": 0.2221744753420353,
|
| 358 |
+
"rewards/reward_fn/mean": 0.5008320093154908,
|
| 359 |
+
"rewards/reward_fn/std": 0.22217447832226753,
|
| 360 |
+
"step": 65,
|
| 361 |
+
"step_time": 22.69940989719953
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"clip_ratio/high_max": 0.0,
|
| 365 |
+
"clip_ratio/high_mean": 0.0,
|
| 366 |
+
"clip_ratio/low_mean": 0.0,
|
| 367 |
+
"clip_ratio/low_min": 0.0,
|
| 368 |
+
"clip_ratio/region_mean": 0.0,
|
| 369 |
+
"completions/clipped_ratio": 0.05,
|
| 370 |
+
"completions/max_length": 149.4,
|
| 371 |
+
"completions/max_terminated_length": 141.4,
|
| 372 |
+
"completions/mean_length": 82.125,
|
| 373 |
+
"completions/mean_terminated_length": 76.2107162475586,
|
| 374 |
+
"completions/min_length": 34.8,
|
| 375 |
+
"completions/min_terminated_length": 34.8,
|
| 376 |
+
"entropy": 0.2419425747357309,
|
| 377 |
+
"epoch": 0.003017241379310345,
|
| 378 |
+
"frac_reward_zero_std": 0.3,
|
| 379 |
+
"grad_norm": 0.5859375,
|
| 380 |
+
"learning_rate": 4.577586206896552e-06,
|
| 381 |
+
"loss": -0.06608393192291259,
|
| 382 |
+
"num_tokens": 162726.0,
|
| 383 |
+
"reward": 0.35277000069618225,
|
| 384 |
+
"reward_std": 0.37774557205848397,
|
| 385 |
+
"rewards/reward_fn/mean": 0.35277000069618225,
|
| 386 |
+
"rewards/reward_fn/std": 0.37774557354860006,
|
| 387 |
+
"step": 70,
|
| 388 |
+
"step_time": 24.30778650340026
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"clip_ratio/high_max": 0.0,
|
| 392 |
+
"clip_ratio/high_mean": 0.0,
|
| 393 |
+
"clip_ratio/low_mean": 0.0,
|
| 394 |
+
"clip_ratio/low_min": 0.0,
|
| 395 |
+
"clip_ratio/region_mean": 0.0,
|
| 396 |
+
"completions/clipped_ratio": 0.025,
|
| 397 |
+
"completions/max_length": 137.6,
|
| 398 |
+
"completions/max_terminated_length": 135.6,
|
| 399 |
+
"completions/mean_length": 77.55,
|
| 400 |
+
"completions/mean_terminated_length": 74.46785888671874,
|
| 401 |
+
"completions/min_length": 41.8,
|
| 402 |
+
"completions/min_terminated_length": 41.8,
|
| 403 |
+
"entropy": 0.35793030727654696,
|
| 404 |
+
"epoch": 0.003232758620689655,
|
| 405 |
+
"frac_reward_zero_std": 0.1,
|
| 406 |
+
"grad_norm": 0.63671875,
|
| 407 |
+
"learning_rate": 4.53448275862069e-06,
|
| 408 |
+
"loss": 0.10921311378479004,
|
| 409 |
+
"num_tokens": 175364.0,
|
| 410 |
+
"reward": 0.15934600830078124,
|
| 411 |
+
"reward_std": 0.5374872148036957,
|
| 412 |
+
"rewards/reward_fn/mean": 0.15934600830078124,
|
| 413 |
+
"rewards/reward_fn/std": 0.5374872386455536,
|
| 414 |
+
"step": 75,
|
| 415 |
+
"step_time": 23.099894465200123
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"clip_ratio/high_max": 0.0,
|
| 419 |
+
"clip_ratio/high_mean": 0.0,
|
| 420 |
+
"clip_ratio/low_mean": 0.0,
|
| 421 |
+
"clip_ratio/low_min": 0.0,
|
| 422 |
+
"clip_ratio/region_mean": 0.0,
|
| 423 |
+
"completions/clipped_ratio": 0.0,
|
| 424 |
+
"completions/max_length": 95.6,
|
| 425 |
+
"completions/max_terminated_length": 95.6,
|
| 426 |
+
"completions/mean_length": 53.175,
|
| 427 |
+
"completions/mean_terminated_length": 53.175,
|
| 428 |
+
"completions/min_length": 15.8,
|
| 429 |
+
"completions/min_terminated_length": 15.8,
|
| 430 |
+
"entropy": 0.2758270605234429,
|
| 431 |
+
"epoch": 0.0034482758620689655,
|
| 432 |
+
"frac_reward_zero_std": 0.3,
|
| 433 |
+
"grad_norm": 0.443359375,
|
| 434 |
+
"learning_rate": 4.4913793103448275e-06,
|
| 435 |
+
"loss": 0.11924625635147094,
|
| 436 |
+
"num_tokens": 186407.0,
|
| 437 |
+
"reward": 0.6774819850921631,
|
| 438 |
+
"reward_std": 0.11054287778679281,
|
| 439 |
+
"rewards/reward_fn/mean": 0.6774819850921631,
|
| 440 |
+
"rewards/reward_fn/std": 0.11054288879968226,
|
| 441 |
+
"step": 80,
|
| 442 |
+
"step_time": 17.348075130399685
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"clip_ratio/high_max": 0.0,
|
| 446 |
+
"clip_ratio/high_mean": 0.0,
|
| 447 |
+
"clip_ratio/low_mean": 0.0,
|
| 448 |
+
"clip_ratio/low_min": 0.0,
|
| 449 |
+
"clip_ratio/region_mean": 0.0,
|
| 450 |
+
"completions/clipped_ratio": 0.0,
|
| 451 |
+
"completions/max_length": 137.6,
|
| 452 |
+
"completions/max_terminated_length": 137.6,
|
| 453 |
+
"completions/mean_length": 68.25,
|
| 454 |
+
"completions/mean_terminated_length": 68.25,
|
| 455 |
+
"completions/min_length": 30.0,
|
| 456 |
+
"completions/min_terminated_length": 30.0,
|
| 457 |
+
"entropy": 0.2696234828326851,
|
| 458 |
+
"epoch": 0.003663793103448276,
|
| 459 |
+
"frac_reward_zero_std": 0.1,
|
| 460 |
+
"grad_norm": 0.287109375,
|
| 461 |
+
"learning_rate": 4.4482758620689656e-06,
|
| 462 |
+
"loss": -0.03822631537914276,
|
| 463 |
+
"num_tokens": 198241.0,
|
| 464 |
+
"reward": 0.5305620029568672,
|
| 465 |
+
"reward_std": 0.4029154841089621,
|
| 466 |
+
"rewards/reward_fn/mean": 0.5305620029568672,
|
| 467 |
+
"rewards/reward_fn/std": 0.4029154877178371,
|
| 468 |
+
"step": 85,
|
| 469 |
+
"step_time": 22.834117503599828
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"clip_ratio/high_max": 0.0,
|
| 473 |
+
"clip_ratio/high_mean": 0.0,
|
| 474 |
+
"clip_ratio/low_mean": 0.0,
|
| 475 |
+
"clip_ratio/low_min": 0.0,
|
| 476 |
+
"clip_ratio/region_mean": 0.0,
|
| 477 |
+
"completions/clipped_ratio": 0.0,
|
| 478 |
+
"completions/max_length": 131.2,
|
| 479 |
+
"completions/max_terminated_length": 131.2,
|
| 480 |
+
"completions/mean_length": 77.45,
|
| 481 |
+
"completions/mean_terminated_length": 77.45,
|
| 482 |
+
"completions/min_length": 19.8,
|
| 483 |
+
"completions/min_terminated_length": 19.8,
|
| 484 |
+
"entropy": 0.3707127865403891,
|
| 485 |
+
"epoch": 0.003879310344827586,
|
| 486 |
+
"frac_reward_zero_std": 0.1,
|
| 487 |
+
"grad_norm": 0.50390625,
|
| 488 |
+
"learning_rate": 4.405172413793104e-06,
|
| 489 |
+
"loss": 0.035861659049987796,
|
| 490 |
+
"num_tokens": 210935.0,
|
| 491 |
+
"reward": 0.3616240084171295,
|
| 492 |
+
"reward_std": 0.48762375079095366,
|
| 493 |
+
"rewards/reward_fn/mean": 0.3616240084171295,
|
| 494 |
+
"rewards/reward_fn/std": 0.48762374445796014,
|
| 495 |
+
"step": 90,
|
| 496 |
+
"step_time": 22.48314213299982
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"clip_ratio/high_max": 0.0,
|
| 500 |
+
"clip_ratio/high_mean": 0.0,
|
| 501 |
+
"clip_ratio/low_mean": 0.0,
|
| 502 |
+
"clip_ratio/low_min": 0.0,
|
| 503 |
+
"clip_ratio/region_mean": 0.0,
|
| 504 |
+
"completions/clipped_ratio": 0.075,
|
| 505 |
+
"completions/max_length": 118.2,
|
| 506 |
+
"completions/max_terminated_length": 116.0,
|
| 507 |
+
"completions/mean_length": 66.9,
|
| 508 |
+
"completions/mean_terminated_length": 56.7,
|
| 509 |
+
"completions/min_length": 25.0,
|
| 510 |
+
"completions/min_terminated_length": 25.0,
|
| 511 |
+
"entropy": 0.342819757014513,
|
| 512 |
+
"epoch": 0.004094827586206897,
|
| 513 |
+
"frac_reward_zero_std": 0.2,
|
| 514 |
+
"grad_norm": 0.78125,
|
| 515 |
+
"learning_rate": 4.362068965517242e-06,
|
| 516 |
+
"loss": -0.051665693521499634,
|
| 517 |
+
"num_tokens": 222647.0,
|
| 518 |
+
"reward": 0.46663198471069334,
|
| 519 |
+
"reward_std": 0.3400567059754394,
|
| 520 |
+
"rewards/reward_fn/mean": 0.46663198471069334,
|
| 521 |
+
"rewards/reward_fn/std": 0.34005671155173334,
|
| 522 |
+
"step": 95,
|
| 523 |
+
"step_time": 20.31996373819993
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"clip_ratio/high_max": 0.0,
|
| 527 |
+
"clip_ratio/high_mean": 0.0,
|
| 528 |
+
"clip_ratio/low_mean": 0.0,
|
| 529 |
+
"clip_ratio/low_min": 0.0,
|
| 530 |
+
"clip_ratio/region_mean": 0.0,
|
| 531 |
+
"completions/clipped_ratio": 0.0,
|
| 532 |
+
"completions/max_length": 82.2,
|
| 533 |
+
"completions/max_terminated_length": 82.2,
|
| 534 |
+
"completions/mean_length": 39.975,
|
| 535 |
+
"completions/mean_terminated_length": 39.975,
|
| 536 |
+
"completions/min_length": 16.4,
|
| 537 |
+
"completions/min_terminated_length": 16.4,
|
| 538 |
+
"entropy": 0.25842549465596676,
|
| 539 |
+
"epoch": 0.004310344827586207,
|
| 540 |
+
"frac_reward_zero_std": 0.2,
|
| 541 |
+
"grad_norm": 0.64453125,
|
| 542 |
+
"learning_rate": 4.31896551724138e-06,
|
| 543 |
+
"loss": 0.029389530420303345,
|
| 544 |
+
"num_tokens": 233466.0,
|
| 545 |
+
"reward": 0.5307899951934815,
|
| 546 |
+
"reward_std": 0.28180397795513273,
|
| 547 |
+
"rewards/reward_fn/mean": 0.5307899951934815,
|
| 548 |
+
"rewards/reward_fn/std": 0.28180398060940204,
|
| 549 |
+
"step": 100,
|
| 550 |
+
"step_time": 15.753981888000453
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"clip_ratio/high_max": 0.0,
|
| 554 |
+
"clip_ratio/high_mean": 0.0,
|
| 555 |
+
"clip_ratio/low_mean": 0.0,
|
| 556 |
+
"clip_ratio/low_min": 0.0,
|
| 557 |
+
"clip_ratio/region_mean": 0.0,
|
| 558 |
+
"completions/clipped_ratio": 0.0,
|
| 559 |
+
"completions/max_length": 77.8,
|
| 560 |
+
"completions/max_terminated_length": 77.8,
|
| 561 |
+
"completions/mean_length": 49.6,
|
| 562 |
+
"completions/mean_terminated_length": 49.6,
|
| 563 |
+
"completions/min_length": 16.4,
|
| 564 |
+
"completions/min_terminated_length": 16.4,
|
| 565 |
+
"entropy": 0.27716706816572695,
|
| 566 |
+
"epoch": 0.004525862068965518,
|
| 567 |
+
"frac_reward_zero_std": 0.6,
|
| 568 |
+
"grad_norm": 0.51953125,
|
| 569 |
+
"learning_rate": 4.275862068965518e-06,
|
| 570 |
+
"loss": 0.011613726615905762,
|
| 571 |
+
"num_tokens": 244550.0,
|
| 572 |
+
"reward": 0.6918019950389862,
|
| 573 |
+
"reward_std": 0.10386521331965923,
|
| 574 |
+
"rewards/reward_fn/mean": 0.6918019950389862,
|
| 575 |
+
"rewards/reward_fn/std": 0.10386521825566888,
|
| 576 |
+
"step": 105,
|
| 577 |
+
"step_time": 15.171548829199946
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"clip_ratio/high_max": 0.0,
|
| 581 |
+
"clip_ratio/high_mean": 0.0,
|
| 582 |
+
"clip_ratio/low_mean": 0.0,
|
| 583 |
+
"clip_ratio/low_min": 0.0,
|
| 584 |
+
"clip_ratio/region_mean": 0.0,
|
| 585 |
+
"completions/clipped_ratio": 0.0,
|
| 586 |
+
"completions/max_length": 94.0,
|
| 587 |
+
"completions/max_terminated_length": 94.0,
|
| 588 |
+
"completions/mean_length": 47.1,
|
| 589 |
+
"completions/mean_terminated_length": 47.1,
|
| 590 |
+
"completions/min_length": 16.0,
|
| 591 |
+
"completions/min_terminated_length": 16.0,
|
| 592 |
+
"entropy": 0.1942274721484864,
|
| 593 |
+
"epoch": 0.0047413793103448275,
|
| 594 |
+
"frac_reward_zero_std": 0.3,
|
| 595 |
+
"grad_norm": 0.515625,
|
| 596 |
+
"learning_rate": 4.232758620689656e-06,
|
| 597 |
+
"loss": -0.000987192615866661,
|
| 598 |
+
"num_tokens": 255962.0,
|
| 599 |
+
"reward": 0.1787539839744568,
|
| 600 |
+
"reward_std": 0.5241983281448483,
|
| 601 |
+
"rewards/reward_fn/mean": 0.1787539839744568,
|
| 602 |
+
"rewards/reward_fn/std": 0.5241983408108354,
|
| 603 |
+
"step": 110,
|
| 604 |
+
"step_time": 17.40271624299976
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"clip_ratio/high_max": 0.0,
|
| 608 |
+
"clip_ratio/high_mean": 0.0,
|
| 609 |
+
"clip_ratio/low_mean": 0.0,
|
| 610 |
+
"clip_ratio/low_min": 0.0,
|
| 611 |
+
"clip_ratio/region_mean": 0.0,
|
| 612 |
+
"completions/clipped_ratio": 0.0,
|
| 613 |
+
"completions/max_length": 108.4,
|
| 614 |
+
"completions/max_terminated_length": 108.4,
|
| 615 |
+
"completions/mean_length": 54.025,
|
| 616 |
+
"completions/mean_terminated_length": 54.025,
|
| 617 |
+
"completions/min_length": 26.8,
|
| 618 |
+
"completions/min_terminated_length": 26.8,
|
| 619 |
+
"entropy": 0.21391218816424953,
|
| 620 |
+
"epoch": 0.004956896551724138,
|
| 621 |
+
"frac_reward_zero_std": 0.1,
|
| 622 |
+
"grad_norm": 0.546875,
|
| 623 |
+
"learning_rate": 4.189655172413794e-06,
|
| 624 |
+
"loss": -0.07643186450004577,
|
| 625 |
+
"num_tokens": 267563.0,
|
| 626 |
+
"reward": 0.5654179990291596,
|
| 627 |
+
"reward_std": 0.2869014423340559,
|
| 628 |
+
"rewards/reward_fn/mean": 0.5654179990291596,
|
| 629 |
+
"rewards/reward_fn/std": 0.2869014598429203,
|
| 630 |
+
"step": 115,
|
| 631 |
+
"step_time": 19.255251389799923
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"clip_ratio/high_max": 0.0,
|
| 635 |
+
"clip_ratio/high_mean": 0.0,
|
| 636 |
+
"clip_ratio/low_mean": 0.0,
|
| 637 |
+
"clip_ratio/low_min": 0.0,
|
| 638 |
+
"clip_ratio/region_mean": 0.0,
|
| 639 |
+
"completions/clipped_ratio": 0.025,
|
| 640 |
+
"completions/max_length": 136.0,
|
| 641 |
+
"completions/max_terminated_length": 124.6,
|
| 642 |
+
"completions/mean_length": 80.575,
|
| 643 |
+
"completions/mean_terminated_length": 77.74285888671875,
|
| 644 |
+
"completions/min_length": 33.0,
|
| 645 |
+
"completions/min_terminated_length": 33.0,
|
| 646 |
+
"entropy": 0.31822405869606885,
|
| 647 |
+
"epoch": 0.005172413793103448,
|
| 648 |
+
"frac_reward_zero_std": 0.2,
|
| 649 |
+
"grad_norm": 1.2421875,
|
| 650 |
+
"learning_rate": 4.146551724137932e-06,
|
| 651 |
+
"loss": 0.01645863652229309,
|
| 652 |
+
"num_tokens": 280074.0,
|
| 653 |
+
"reward": 0.06659798473119735,
|
| 654 |
+
"reward_std": 0.5049000799655914,
|
| 655 |
+
"rewards/reward_fn/mean": 0.06659798473119735,
|
| 656 |
+
"rewards/reward_fn/std": 0.5049000844359398,
|
| 657 |
+
"step": 120,
|
| 658 |
+
"step_time": 22.649546912000005
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"clip_ratio/high_max": 0.0,
|
| 662 |
+
"clip_ratio/high_mean": 0.0,
|
| 663 |
+
"clip_ratio/low_mean": 0.0,
|
| 664 |
+
"clip_ratio/low_min": 0.0,
|
| 665 |
+
"clip_ratio/region_mean": 0.0,
|
| 666 |
+
"completions/clipped_ratio": 0.05,
|
| 667 |
+
"completions/max_length": 138.4,
|
| 668 |
+
"completions/max_terminated_length": 136.2,
|
| 669 |
+
"completions/mean_length": 68.85,
|
| 670 |
+
"completions/mean_terminated_length": 61.975,
|
| 671 |
+
"completions/min_length": 25.4,
|
| 672 |
+
"completions/min_terminated_length": 25.4,
|
| 673 |
+
"entropy": 0.33708888398250564,
|
| 674 |
+
"epoch": 0.005387931034482759,
|
| 675 |
+
"frac_reward_zero_std": 0.1,
|
| 676 |
+
"grad_norm": 1.25,
|
| 677 |
+
"learning_rate": 4.103448275862069e-06,
|
| 678 |
+
"loss": 0.16269092559814452,
|
| 679 |
+
"num_tokens": 291760.0,
|
| 680 |
+
"reward": 0.09610399603843689,
|
| 681 |
+
"reward_std": 0.6181229472160339,
|
| 682 |
+
"rewards/reward_fn/mean": 0.09610399603843689,
|
| 683 |
+
"rewards/reward_fn/std": 0.6181229710578918,
|
| 684 |
+
"step": 125,
|
| 685 |
+
"step_time": 22.892032967400247
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"clip_ratio/high_max": 0.0,
|
| 689 |
+
"clip_ratio/high_mean": 0.0,
|
| 690 |
+
"clip_ratio/low_mean": 0.0,
|
| 691 |
+
"clip_ratio/low_min": 0.0,
|
| 692 |
+
"clip_ratio/region_mean": 0.0,
|
| 693 |
+
"completions/clipped_ratio": 0.05,
|
| 694 |
+
"completions/max_length": 121.6,
|
| 695 |
+
"completions/max_terminated_length": 119.6,
|
| 696 |
+
"completions/mean_length": 54.625,
|
| 697 |
+
"completions/mean_terminated_length": 45.516667175292966,
|
| 698 |
+
"completions/min_length": 15.4,
|
| 699 |
+
"completions/min_terminated_length": 15.4,
|
| 700 |
+
"entropy": 0.2316889252353576,
|
| 701 |
+
"epoch": 0.005603448275862069,
|
| 702 |
+
"frac_reward_zero_std": 0.1,
|
| 703 |
+
"grad_norm": 0.78515625,
|
| 704 |
+
"learning_rate": 4.060344827586207e-06,
|
| 705 |
+
"loss": 0.05887578129768371,
|
| 706 |
+
"num_tokens": 302777.0,
|
| 707 |
+
"reward": 0.47827199250459673,
|
| 708 |
+
"reward_std": 0.31257240688428284,
|
| 709 |
+
"rewards/reward_fn/mean": 0.47827199250459673,
|
| 710 |
+
"rewards/reward_fn/std": 0.31257240623235705,
|
| 711 |
+
"step": 130,
|
| 712 |
+
"step_time": 20.756939186200178
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"clip_ratio/high_max": 0.0,
|
| 716 |
+
"clip_ratio/high_mean": 0.0,
|
| 717 |
+
"clip_ratio/low_mean": 0.0,
|
| 718 |
+
"clip_ratio/low_min": 0.0,
|
| 719 |
+
"clip_ratio/region_mean": 0.0,
|
| 720 |
+
"completions/clipped_ratio": 0.0,
|
| 721 |
+
"completions/max_length": 83.4,
|
| 722 |
+
"completions/max_terminated_length": 83.4,
|
| 723 |
+
"completions/mean_length": 43.375,
|
| 724 |
+
"completions/mean_terminated_length": 43.375,
|
| 725 |
+
"completions/min_length": 17.0,
|
| 726 |
+
"completions/min_terminated_length": 17.0,
|
| 727 |
+
"entropy": 0.14220066292909905,
|
| 728 |
+
"epoch": 0.00581896551724138,
|
| 729 |
+
"frac_reward_zero_std": 0.3,
|
| 730 |
+
"grad_norm": 0.361328125,
|
| 731 |
+
"learning_rate": 4.017241379310346e-06,
|
| 732 |
+
"loss": 0.0067549549043178555,
|
| 733 |
+
"num_tokens": 313896.0,
|
| 734 |
+
"reward": 0.5540139883756637,
|
| 735 |
+
"reward_std": 0.3137821853160858,
|
| 736 |
+
"rewards/reward_fn/mean": 0.5540139883756637,
|
| 737 |
+
"rewards/reward_fn/std": 0.31378220971673726,
|
| 738 |
+
"step": 135,
|
| 739 |
+
"step_time": 16.04840490619972
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"clip_ratio/high_max": 0.0,
|
| 743 |
+
"clip_ratio/high_mean": 0.0,
|
| 744 |
+
"clip_ratio/low_mean": 0.0,
|
| 745 |
+
"clip_ratio/low_min": 0.0,
|
| 746 |
+
"clip_ratio/region_mean": 0.0,
|
| 747 |
+
"completions/clipped_ratio": 0.0,
|
| 748 |
+
"completions/max_length": 65.2,
|
| 749 |
+
"completions/max_terminated_length": 65.2,
|
| 750 |
+
"completions/mean_length": 33.825,
|
| 751 |
+
"completions/mean_terminated_length": 33.825,
|
| 752 |
+
"completions/min_length": 15.8,
|
| 753 |
+
"completions/min_terminated_length": 15.8,
|
| 754 |
+
"entropy": 0.16708708838559688,
|
| 755 |
+
"epoch": 0.00603448275862069,
|
| 756 |
+
"frac_reward_zero_std": 0.3,
|
| 757 |
+
"grad_norm": 0.53125,
|
| 758 |
+
"learning_rate": 3.974137931034483e-06,
|
| 759 |
+
"loss": 0.000922510027885437,
|
| 760 |
+
"num_tokens": 324093.0,
|
| 761 |
+
"reward": 0.5665820002555847,
|
| 762 |
+
"reward_std": 0.3455619063111953,
|
| 763 |
+
"rewards/reward_fn/mean": 0.5665820002555847,
|
| 764 |
+
"rewards/reward_fn/std": 0.34556191868614405,
|
| 765 |
+
"step": 140,
|
| 766 |
+
"step_time": 13.251863584200146
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"clip_ratio/high_max": 0.0,
|
| 770 |
+
"clip_ratio/high_mean": 0.0,
|
| 771 |
+
"clip_ratio/low_mean": 0.0,
|
| 772 |
+
"clip_ratio/low_min": 0.0,
|
| 773 |
+
"clip_ratio/region_mean": 0.0,
|
| 774 |
+
"completions/clipped_ratio": 0.1,
|
| 775 |
+
"completions/max_length": 175.8,
|
| 776 |
+
"completions/max_terminated_length": 149.4,
|
| 777 |
+
"completions/mean_length": 76.9,
|
| 778 |
+
"completions/mean_terminated_length": 58.73142929077149,
|
| 779 |
+
"completions/min_length": 25.6,
|
| 780 |
+
"completions/min_terminated_length": 25.6,
|
| 781 |
+
"entropy": 0.2752605409361422,
|
| 782 |
+
"epoch": 0.00625,
|
| 783 |
+
"frac_reward_zero_std": 0.0,
|
| 784 |
+
"grad_norm": 0.62109375,
|
| 785 |
+
"learning_rate": 3.931034482758621e-06,
|
| 786 |
+
"loss": -0.012939244508743286,
|
| 787 |
+
"num_tokens": 336493.0,
|
| 788 |
+
"reward": 0.03320800065994263,
|
| 789 |
+
"reward_std": 0.6782135218381882,
|
| 790 |
+
"rewards/reward_fn/mean": 0.03320800065994263,
|
| 791 |
+
"rewards/reward_fn/std": 0.6782135486602783,
|
| 792 |
+
"step": 145,
|
| 793 |
+
"step_time": 27.717435833599847
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"clip_ratio/high_max": 0.0,
|
| 797 |
+
"clip_ratio/high_mean": 0.0,
|
| 798 |
+
"clip_ratio/low_mean": 0.0,
|
| 799 |
+
"clip_ratio/low_min": 0.0,
|
| 800 |
+
"clip_ratio/region_mean": 0.0,
|
| 801 |
+
"completions/clipped_ratio": 0.05,
|
| 802 |
+
"completions/max_length": 159.6,
|
| 803 |
+
"completions/max_terminated_length": 159.4,
|
| 804 |
+
"completions/mean_length": 86.4,
|
| 805 |
+
"completions/mean_terminated_length": 78.2,
|
| 806 |
+
"completions/min_length": 27.0,
|
| 807 |
+
"completions/min_terminated_length": 27.0,
|
| 808 |
+
"entropy": 0.3023837681859732,
|
| 809 |
+
"epoch": 0.00646551724137931,
|
| 810 |
+
"frac_reward_zero_std": 0.1,
|
| 811 |
+
"grad_norm": 0.25,
|
| 812 |
+
"learning_rate": 3.887931034482759e-06,
|
| 813 |
+
"loss": -0.004035864770412445,
|
| 814 |
+
"num_tokens": 349213.0,
|
| 815 |
+
"reward": 0.2794659972190857,
|
| 816 |
+
"reward_std": 0.33909483794122935,
|
| 817 |
+
"rewards/reward_fn/mean": 0.2794659972190857,
|
| 818 |
+
"rewards/reward_fn/std": 0.3390948371961713,
|
| 819 |
+
"step": 150,
|
| 820 |
+
"step_time": 25.6254667358
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"clip_ratio/high_max": 0.0,
|
| 824 |
+
"clip_ratio/high_mean": 0.0,
|
| 825 |
+
"clip_ratio/low_mean": 0.0,
|
| 826 |
+
"clip_ratio/low_min": 0.0,
|
| 827 |
+
"clip_ratio/region_mean": 0.0,
|
| 828 |
+
"completions/clipped_ratio": 0.025,
|
| 829 |
+
"completions/max_length": 120.8,
|
| 830 |
+
"completions/max_terminated_length": 117.2,
|
| 831 |
+
"completions/mean_length": 76.35,
|
| 832 |
+
"completions/mean_terminated_length": 74.68571472167969,
|
| 833 |
+
"completions/min_length": 42.2,
|
| 834 |
+
"completions/min_terminated_length": 42.2,
|
| 835 |
+
"entropy": 0.2444805505729164,
|
| 836 |
+
"epoch": 0.006681034482758621,
|
| 837 |
+
"frac_reward_zero_std": 0.3,
|
| 838 |
+
"grad_norm": 0.59375,
|
| 839 |
+
"learning_rate": 3.844827586206897e-06,
|
| 840 |
+
"loss": 0.06990929841995239,
|
| 841 |
+
"num_tokens": 361243.0,
|
| 842 |
+
"reward": 0.31372600197792055,
|
| 843 |
+
"reward_std": 0.16792407557368277,
|
| 844 |
+
"rewards/reward_fn/mean": 0.31372600197792055,
|
| 845 |
+
"rewards/reward_fn/std": 0.16792407706379892,
|
| 846 |
+
"step": 155,
|
| 847 |
+
"step_time": 20.657237647200326
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"clip_ratio/high_max": 0.0,
|
| 851 |
+
"clip_ratio/high_mean": 0.0,
|
| 852 |
+
"clip_ratio/low_mean": 0.0,
|
| 853 |
+
"clip_ratio/low_min": 0.0,
|
| 854 |
+
"clip_ratio/region_mean": 0.0,
|
| 855 |
+
"completions/clipped_ratio": 0.025,
|
| 856 |
+
"completions/max_length": 86.2,
|
| 857 |
+
"completions/max_terminated_length": 78.2,
|
| 858 |
+
"completions/mean_length": 57.375,
|
| 859 |
+
"completions/mean_terminated_length": 54.985714721679685,
|
| 860 |
+
"completions/min_length": 36.2,
|
| 861 |
+
"completions/min_terminated_length": 36.2,
|
| 862 |
+
"entropy": 0.2041449310956523,
|
| 863 |
+
"epoch": 0.006896551724137931,
|
| 864 |
+
"frac_reward_zero_std": 0.3,
|
| 865 |
+
"grad_norm": 0.146484375,
|
| 866 |
+
"learning_rate": 3.801724137931035e-06,
|
| 867 |
+
"loss": -0.003799402713775635,
|
| 868 |
+
"num_tokens": 372606.0,
|
| 869 |
+
"reward": 0.6000540047883988,
|
| 870 |
+
"reward_std": 0.14438311448320745,
|
| 871 |
+
"rewards/reward_fn/mean": 0.6000540047883988,
|
| 872 |
+
"rewards/reward_fn/std": 0.1443831158336252,
|
| 873 |
+
"step": 160,
|
| 874 |
+
"step_time": 15.94448771479947
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"clip_ratio/high_max": 0.0,
|
| 878 |
+
"clip_ratio/high_mean": 0.0,
|
| 879 |
+
"clip_ratio/low_mean": 0.0,
|
| 880 |
+
"clip_ratio/low_min": 0.0,
|
| 881 |
+
"clip_ratio/region_mean": 0.0,
|
| 882 |
+
"completions/clipped_ratio": 0.1,
|
| 883 |
+
"completions/max_length": 196.6,
|
| 884 |
+
"completions/max_terminated_length": 173.8,
|
| 885 |
+
"completions/mean_length": 114.125,
|
| 886 |
+
"completions/mean_terminated_length": 102.28333435058593,
|
| 887 |
+
"completions/min_length": 40.8,
|
| 888 |
+
"completions/min_terminated_length": 40.8,
|
| 889 |
+
"entropy": 0.3578500231727958,
|
| 890 |
+
"epoch": 0.007112068965517242,
|
| 891 |
+
"frac_reward_zero_std": 0.0,
|
| 892 |
+
"grad_norm": 0.515625,
|
| 893 |
+
"learning_rate": 3.7586206896551727e-06,
|
| 894 |
+
"loss": 0.05463705062866211,
|
| 895 |
+
"num_tokens": 386899.0,
|
| 896 |
+
"reward": 0.29335000514984133,
|
| 897 |
+
"reward_std": 0.6011195421218872,
|
| 898 |
+
"rewards/reward_fn/mean": 0.29335000514984133,
|
| 899 |
+
"rewards/reward_fn/std": 0.6011195570230484,
|
| 900 |
+
"step": 165,
|
| 901 |
+
"step_time": 30.583657701599986
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"clip_ratio/high_max": 0.0,
|
| 905 |
+
"clip_ratio/high_mean": 0.0,
|
| 906 |
+
"clip_ratio/low_mean": 0.0,
|
| 907 |
+
"clip_ratio/low_min": 0.0,
|
| 908 |
+
"clip_ratio/region_mean": 0.0,
|
| 909 |
+
"completions/clipped_ratio": 0.025,
|
| 910 |
+
"completions/max_length": 120.6,
|
| 911 |
+
"completions/max_terminated_length": 105.4,
|
| 912 |
+
"completions/mean_length": 64.775,
|
| 913 |
+
"completions/mean_terminated_length": 60.79285736083985,
|
| 914 |
+
"completions/min_length": 27.2,
|
| 915 |
+
"completions/min_terminated_length": 27.2,
|
| 916 |
+
"entropy": 0.3172524501627777,
|
| 917 |
+
"epoch": 0.007327586206896552,
|
| 918 |
+
"frac_reward_zero_std": 0.2,
|
| 919 |
+
"grad_norm": 0.380859375,
|
| 920 |
+
"learning_rate": 3.7155172413793107e-06,
|
| 921 |
+
"loss": -0.04064536094665527,
|
| 922 |
+
"num_tokens": 398878.0,
|
| 923 |
+
"reward": 0.2578879952430725,
|
| 924 |
+
"reward_std": 0.4871767422184348,
|
| 925 |
+
"rewards/reward_fn/mean": 0.2578879952430725,
|
| 926 |
+
"rewards/reward_fn/std": 0.4871767588891089,
|
| 927 |
+
"step": 170,
|
| 928 |
+
"step_time": 20.961449378399994
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"clip_ratio/high_max": 0.0,
|
| 932 |
+
"clip_ratio/high_mean": 0.0,
|
| 933 |
+
"clip_ratio/low_mean": 0.0,
|
| 934 |
+
"clip_ratio/low_min": 0.0,
|
| 935 |
+
"clip_ratio/region_mean": 0.0,
|
| 936 |
+
"completions/clipped_ratio": 0.0,
|
| 937 |
+
"completions/max_length": 86.0,
|
| 938 |
+
"completions/max_terminated_length": 86.0,
|
| 939 |
+
"completions/mean_length": 49.1,
|
| 940 |
+
"completions/mean_terminated_length": 49.1,
|
| 941 |
+
"completions/min_length": 25.4,
|
| 942 |
+
"completions/min_terminated_length": 25.4,
|
| 943 |
+
"entropy": 0.28907319878344423,
|
| 944 |
+
"epoch": 0.007543103448275862,
|
| 945 |
+
"frac_reward_zero_std": 0.3,
|
| 946 |
+
"grad_norm": 0.63671875,
|
| 947 |
+
"learning_rate": 3.6724137931034487e-06,
|
| 948 |
+
"loss": 0.07596501111984252,
|
| 949 |
+
"num_tokens": 409894.0,
|
| 950 |
+
"reward": 0.2182680130004883,
|
| 951 |
+
"reward_std": 0.2655696720117703,
|
| 952 |
+
"rewards/reward_fn/mean": 0.2182680130004883,
|
| 953 |
+
"rewards/reward_fn/std": 0.26556967524811625,
|
| 954 |
+
"step": 175,
|
| 955 |
+
"step_time": 16.00376275539984
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"clip_ratio/high_max": 0.0,
|
| 959 |
+
"clip_ratio/high_mean": 0.0,
|
| 960 |
+
"clip_ratio/low_mean": 0.0,
|
| 961 |
+
"clip_ratio/low_min": 0.0,
|
| 962 |
+
"clip_ratio/region_mean": 0.0,
|
| 963 |
+
"completions/clipped_ratio": 0.075,
|
| 964 |
+
"completions/max_length": 119.0,
|
| 965 |
+
"completions/max_terminated_length": 118.0,
|
| 966 |
+
"completions/mean_length": 75.025,
|
| 967 |
+
"completions/mean_terminated_length": 60.685000610351565,
|
| 968 |
+
"completions/min_length": 30.4,
|
| 969 |
+
"completions/min_terminated_length": 30.4,
|
| 970 |
+
"entropy": 0.2300539159856271,
|
| 971 |
+
"epoch": 0.007758620689655172,
|
| 972 |
+
"frac_reward_zero_std": 0.4,
|
| 973 |
+
"grad_norm": 0.33203125,
|
| 974 |
+
"learning_rate": 3.6293103448275863e-06,
|
| 975 |
+
"loss": 0.0008989959955215454,
|
| 976 |
+
"num_tokens": 422115.0,
|
| 977 |
+
"reward": 0.2748300015926361,
|
| 978 |
+
"reward_std": 0.4872116835438646,
|
| 979 |
+
"rewards/reward_fn/mean": 0.2748300015926361,
|
| 980 |
+
"rewards/reward_fn/std": 0.487211688212119,
|
| 981 |
+
"step": 180,
|
| 982 |
+
"step_time": 20.472563396999977
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"clip_ratio/high_max": 0.0,
|
| 986 |
+
"clip_ratio/high_mean": 0.0,
|
| 987 |
+
"clip_ratio/low_mean": 0.0,
|
| 988 |
+
"clip_ratio/low_min": 0.0,
|
| 989 |
+
"clip_ratio/region_mean": 0.0,
|
| 990 |
+
"completions/clipped_ratio": 0.025,
|
| 991 |
+
"completions/max_length": 140.8,
|
| 992 |
+
"completions/max_terminated_length": 130.4,
|
| 993 |
+
"completions/mean_length": 77.75,
|
| 994 |
+
"completions/mean_terminated_length": 73.95714416503907,
|
| 995 |
+
"completions/min_length": 28.4,
|
| 996 |
+
"completions/min_terminated_length": 28.4,
|
| 997 |
+
"entropy": 0.2450535054318607,
|
| 998 |
+
"epoch": 0.007974137931034483,
|
| 999 |
+
"frac_reward_zero_std": 0.3,
|
| 1000 |
+
"grad_norm": 0.0,
|
| 1001 |
+
"learning_rate": 3.5862068965517243e-06,
|
| 1002 |
+
"loss": 0.041662493348121644,
|
| 1003 |
+
"num_tokens": 434661.0,
|
| 1004 |
+
"reward": 0.052784007787704465,
|
| 1005 |
+
"reward_std": 0.2650376170873642,
|
| 1006 |
+
"rewards/reward_fn/mean": 0.052784007787704465,
|
| 1007 |
+
"rewards/reward_fn/std": 0.26503762900829314,
|
| 1008 |
+
"step": 185,
|
| 1009 |
+
"step_time": 23.318358728199563
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"clip_ratio/high_max": 0.0,
|
| 1013 |
+
"clip_ratio/high_mean": 0.0,
|
| 1014 |
+
"clip_ratio/low_mean": 0.0,
|
| 1015 |
+
"clip_ratio/low_min": 0.0,
|
| 1016 |
+
"clip_ratio/region_mean": 0.0,
|
| 1017 |
+
"completions/clipped_ratio": 0.05,
|
| 1018 |
+
"completions/max_length": 195.6,
|
| 1019 |
+
"completions/max_terminated_length": 192.8,
|
| 1020 |
+
"completions/mean_length": 89.9,
|
| 1021 |
+
"completions/mean_terminated_length": 84.275,
|
| 1022 |
+
"completions/min_length": 26.2,
|
| 1023 |
+
"completions/min_terminated_length": 26.2,
|
| 1024 |
+
"entropy": 0.3329991169273853,
|
| 1025 |
+
"epoch": 0.008189655172413794,
|
| 1026 |
+
"frac_reward_zero_std": 0.3,
|
| 1027 |
+
"grad_norm": 0.3515625,
|
| 1028 |
+
"learning_rate": 3.5431034482758623e-06,
|
| 1029 |
+
"loss": 0.014610832929611206,
|
| 1030 |
+
"num_tokens": 447801.0,
|
| 1031 |
+
"reward": 0.449290007352829,
|
| 1032 |
+
"reward_std": 0.2617717768996954,
|
| 1033 |
+
"rewards/reward_fn/mean": 0.449290007352829,
|
| 1034 |
+
"rewards/reward_fn/std": 0.26177178248763083,
|
| 1035 |
+
"step": 190,
|
| 1036 |
+
"step_time": 30.514870558599977
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"clip_ratio/high_max": 0.0,
|
| 1040 |
+
"clip_ratio/high_mean": 0.0,
|
| 1041 |
+
"clip_ratio/low_mean": 0.0,
|
| 1042 |
+
"clip_ratio/low_min": 0.0,
|
| 1043 |
+
"clip_ratio/region_mean": 0.0,
|
| 1044 |
+
"completions/clipped_ratio": 0.075,
|
| 1045 |
+
"completions/max_length": 174.6,
|
| 1046 |
+
"completions/max_terminated_length": 167.2,
|
| 1047 |
+
"completions/mean_length": 102.6,
|
| 1048 |
+
"completions/mean_terminated_length": 93.59166717529297,
|
| 1049 |
+
"completions/min_length": 33.8,
|
| 1050 |
+
"completions/min_terminated_length": 33.8,
|
| 1051 |
+
"entropy": 0.27610186783713286,
|
| 1052 |
+
"epoch": 0.008405172413793103,
|
| 1053 |
+
"frac_reward_zero_std": 0.4,
|
| 1054 |
+
"grad_norm": 0.33203125,
|
| 1055 |
+
"learning_rate": 3.5e-06,
|
| 1056 |
+
"loss": 0.014060372114181518,
|
| 1057 |
+
"num_tokens": 461297.0,
|
| 1058 |
+
"reward": 0.18859200328588485,
|
| 1059 |
+
"reward_std": 0.20913713360205294,
|
| 1060 |
+
"rewards/reward_fn/mean": 0.18859200328588485,
|
| 1061 |
+
"rewards/reward_fn/std": 0.20913713917834684,
|
| 1062 |
+
"step": 195,
|
| 1063 |
+
"step_time": 27.840963185400142
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"clip_ratio/high_max": 0.0,
|
| 1067 |
+
"clip_ratio/high_mean": 0.0,
|
| 1068 |
+
"clip_ratio/low_mean": 0.0,
|
| 1069 |
+
"clip_ratio/low_min": 0.0,
|
| 1070 |
+
"clip_ratio/region_mean": 0.0,
|
| 1071 |
+
"completions/clipped_ratio": 0.1,
|
| 1072 |
+
"completions/max_length": 145.6,
|
| 1073 |
+
"completions/max_terminated_length": 108.0,
|
| 1074 |
+
"completions/mean_length": 63.075,
|
| 1075 |
+
"completions/mean_terminated_length": 43.47500076293945,
|
| 1076 |
+
"completions/min_length": 16.4,
|
| 1077 |
+
"completions/min_terminated_length": 16.4,
|
| 1078 |
+
"entropy": 0.20764350929530337,
|
| 1079 |
+
"epoch": 0.008620689655172414,
|
| 1080 |
+
"frac_reward_zero_std": 0.3,
|
| 1081 |
+
"grad_norm": 0.2177734375,
|
| 1082 |
+
"learning_rate": 3.4568965517241384e-06,
|
| 1083 |
+
"loss": 0.01216265857219696,
|
| 1084 |
+
"num_tokens": 473228.0,
|
| 1085 |
+
"reward": 0.4162260115146637,
|
| 1086 |
+
"reward_std": 0.5489736057817935,
|
| 1087 |
+
"rewards/reward_fn/mean": 0.4162260115146637,
|
| 1088 |
+
"rewards/reward_fn/std": 0.5489736313116736,
|
| 1089 |
+
"step": 200,
|
| 1090 |
+
"step_time": 24.10546086580016
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"clip_ratio/high_max": 0.0,
|
| 1094 |
+
"clip_ratio/high_mean": 0.0,
|
| 1095 |
+
"clip_ratio/low_mean": 0.0,
|
| 1096 |
+
"clip_ratio/low_min": 0.0,
|
| 1097 |
+
"clip_ratio/region_mean": 0.0,
|
| 1098 |
+
"completions/clipped_ratio": 0.0,
|
| 1099 |
+
"completions/max_length": 55.6,
|
| 1100 |
+
"completions/max_terminated_length": 55.6,
|
| 1101 |
+
"completions/mean_length": 23.825,
|
| 1102 |
+
"completions/mean_terminated_length": 23.825,
|
| 1103 |
+
"completions/min_length": 16.0,
|
| 1104 |
+
"completions/min_terminated_length": 16.0,
|
| 1105 |
+
"entropy": 0.13016226456093136,
|
| 1106 |
+
"epoch": 0.008836206896551724,
|
| 1107 |
+
"frac_reward_zero_std": 0.3,
|
| 1108 |
+
"grad_norm": 2.59375,
|
| 1109 |
+
"learning_rate": 3.4137931034482764e-06,
|
| 1110 |
+
"loss": 0.10798157453536987,
|
| 1111 |
+
"num_tokens": 483053.0,
|
| 1112 |
+
"reward": 0.7742319941520691,
|
| 1113 |
+
"reward_std": 0.06249405808048323,
|
| 1114 |
+
"rewards/reward_fn/mean": 0.7742319941520691,
|
| 1115 |
+
"rewards/reward_fn/std": 0.06249405494891107,
|
| 1116 |
+
"step": 205,
|
| 1117 |
+
"step_time": 12.142383367000003
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"clip_ratio/high_max": 0.0,
|
| 1121 |
+
"clip_ratio/high_mean": 0.0,
|
| 1122 |
+
"clip_ratio/low_mean": 0.0,
|
| 1123 |
+
"clip_ratio/low_min": 0.0,
|
| 1124 |
+
"clip_ratio/region_mean": 0.0,
|
| 1125 |
+
"completions/clipped_ratio": 0.0,
|
| 1126 |
+
"completions/max_length": 104.8,
|
| 1127 |
+
"completions/max_terminated_length": 104.8,
|
| 1128 |
+
"completions/mean_length": 50.125,
|
| 1129 |
+
"completions/mean_terminated_length": 50.125,
|
| 1130 |
+
"completions/min_length": 22.4,
|
| 1131 |
+
"completions/min_terminated_length": 22.4,
|
| 1132 |
+
"entropy": 0.3128367411205545,
|
| 1133 |
+
"epoch": 0.009051724137931035,
|
| 1134 |
+
"frac_reward_zero_std": 0.2,
|
| 1135 |
+
"grad_norm": 0.57421875,
|
| 1136 |
+
"learning_rate": 3.370689655172414e-06,
|
| 1137 |
+
"loss": -0.012002941966056824,
|
| 1138 |
+
"num_tokens": 494054.0,
|
| 1139 |
+
"reward": 0.31811599135398866,
|
| 1140 |
+
"reward_std": 0.6536980773322284,
|
| 1141 |
+
"rewards/reward_fn/mean": 0.31811599135398866,
|
| 1142 |
+
"rewards/reward_fn/std": 0.6536980641074479,
|
| 1143 |
+
"step": 210,
|
| 1144 |
+
"step_time": 18.774766382600138
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"clip_ratio/high_max": 0.0,
|
| 1148 |
+
"clip_ratio/high_mean": 0.0,
|
| 1149 |
+
"clip_ratio/low_mean": 0.0,
|
| 1150 |
+
"clip_ratio/low_min": 0.0,
|
| 1151 |
+
"clip_ratio/region_mean": 0.0,
|
| 1152 |
+
"completions/clipped_ratio": 0.0,
|
| 1153 |
+
"completions/max_length": 79.8,
|
| 1154 |
+
"completions/max_terminated_length": 79.8,
|
| 1155 |
+
"completions/mean_length": 54.7,
|
| 1156 |
+
"completions/mean_terminated_length": 54.7,
|
| 1157 |
+
"completions/min_length": 38.4,
|
| 1158 |
+
"completions/min_terminated_length": 38.4,
|
| 1159 |
+
"entropy": 0.18855432691198076,
|
| 1160 |
+
"epoch": 0.009267241379310344,
|
| 1161 |
+
"frac_reward_zero_std": 0.3,
|
| 1162 |
+
"grad_norm": 0.376953125,
|
| 1163 |
+
"learning_rate": 3.327586206896552e-06,
|
| 1164 |
+
"loss": -0.01920826584100723,
|
| 1165 |
+
"num_tokens": 505226.0,
|
| 1166 |
+
"reward": 0.5115379929542542,
|
| 1167 |
+
"reward_std": 0.15735746768768877,
|
| 1168 |
+
"rewards/reward_fn/mean": 0.5115379929542542,
|
| 1169 |
+
"rewards/reward_fn/std": 0.1573574845213443,
|
| 1170 |
+
"step": 215,
|
| 1171 |
+
"step_time": 15.221036481800548
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"clip_ratio/high_max": 0.0,
|
| 1175 |
+
"clip_ratio/high_mean": 0.0,
|
| 1176 |
+
"clip_ratio/low_mean": 0.0,
|
| 1177 |
+
"clip_ratio/low_min": 0.0,
|
| 1178 |
+
"clip_ratio/region_mean": 0.0,
|
| 1179 |
+
"completions/clipped_ratio": 0.0,
|
| 1180 |
+
"completions/max_length": 126.2,
|
| 1181 |
+
"completions/max_terminated_length": 126.2,
|
| 1182 |
+
"completions/mean_length": 59.2,
|
| 1183 |
+
"completions/mean_terminated_length": 59.2,
|
| 1184 |
+
"completions/min_length": 16.2,
|
| 1185 |
+
"completions/min_terminated_length": 16.2,
|
| 1186 |
+
"entropy": 0.31272616144269705,
|
| 1187 |
+
"epoch": 0.009482758620689655,
|
| 1188 |
+
"frac_reward_zero_std": 0.1,
|
| 1189 |
+
"grad_norm": 0.7578125,
|
| 1190 |
+
"learning_rate": 3.28448275862069e-06,
|
| 1191 |
+
"loss": 0.13071467876434326,
|
| 1192 |
+
"num_tokens": 516730.0,
|
| 1193 |
+
"reward": 0.6471440196037292,
|
| 1194 |
+
"reward_std": 0.15736471198033541,
|
| 1195 |
+
"rewards/reward_fn/mean": 0.6471440196037292,
|
| 1196 |
+
"rewards/reward_fn/std": 0.15736472072312607,
|
| 1197 |
+
"step": 220,
|
| 1198 |
+
"step_time": 21.683129164199453
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"clip_ratio/high_max": 0.0,
|
| 1202 |
+
"clip_ratio/high_mean": 0.0,
|
| 1203 |
+
"clip_ratio/low_mean": 0.0,
|
| 1204 |
+
"clip_ratio/low_min": 0.0,
|
| 1205 |
+
"clip_ratio/region_mean": 0.0,
|
| 1206 |
+
"completions/clipped_ratio": 0.0,
|
| 1207 |
+
"completions/max_length": 75.6,
|
| 1208 |
+
"completions/max_terminated_length": 75.6,
|
| 1209 |
+
"completions/mean_length": 41.075,
|
| 1210 |
+
"completions/mean_terminated_length": 41.075,
|
| 1211 |
+
"completions/min_length": 21.2,
|
| 1212 |
+
"completions/min_terminated_length": 21.2,
|
| 1213 |
+
"entropy": 0.17901114720734768,
|
| 1214 |
+
"epoch": 0.009698275862068966,
|
| 1215 |
+
"frac_reward_zero_std": 0.2,
|
| 1216 |
+
"grad_norm": 0.390625,
|
| 1217 |
+
"learning_rate": 3.2413793103448277e-06,
|
| 1218 |
+
"loss": 0.0042393475770950316,
|
| 1219 |
+
"num_tokens": 527637.0,
|
| 1220 |
+
"reward": 0.5594579905271531,
|
| 1221 |
+
"reward_std": 0.23395470455288886,
|
| 1222 |
+
"rewards/reward_fn/mean": 0.5594579905271531,
|
| 1223 |
+
"rewards/reward_fn/std": 0.23395471236435697,
|
| 1224 |
+
"step": 225,
|
| 1225 |
+
"step_time": 14.985401752000143
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"clip_ratio/high_max": 0.0,
|
| 1229 |
+
"clip_ratio/high_mean": 0.0,
|
| 1230 |
+
"clip_ratio/low_mean": 0.0,
|
| 1231 |
+
"clip_ratio/low_min": 0.0,
|
| 1232 |
+
"clip_ratio/region_mean": 0.0,
|
| 1233 |
+
"completions/clipped_ratio": 0.025,
|
| 1234 |
+
"completions/max_length": 145.0,
|
| 1235 |
+
"completions/max_terminated_length": 144.0,
|
| 1236 |
+
"completions/mean_length": 77.85,
|
| 1237 |
+
"completions/mean_terminated_length": 73.91785736083985,
|
| 1238 |
+
"completions/min_length": 29.8,
|
| 1239 |
+
"completions/min_terminated_length": 29.8,
|
| 1240 |
+
"entropy": 0.28966000496875494,
|
| 1241 |
+
"epoch": 0.009913793103448277,
|
| 1242 |
+
"frac_reward_zero_std": 0.3,
|
| 1243 |
+
"grad_norm": 0.294921875,
|
| 1244 |
+
"learning_rate": 3.1982758620689657e-06,
|
| 1245 |
+
"loss": -0.03884480595588684,
|
| 1246 |
+
"num_tokens": 539999.0,
|
| 1247 |
+
"reward": 0.2210400015115738,
|
| 1248 |
+
"reward_std": 0.45898649455048146,
|
| 1249 |
+
"rewards/reward_fn/mean": 0.2210400015115738,
|
| 1250 |
+
"rewards/reward_fn/std": 0.458986507600639,
|
| 1251 |
+
"step": 230,
|
| 1252 |
+
"step_time": 23.7599374403997
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"clip_ratio/high_max": 0.0,
|
| 1256 |
+
"clip_ratio/high_mean": 0.0,
|
| 1257 |
+
"clip_ratio/low_mean": 0.0,
|
| 1258 |
+
"clip_ratio/low_min": 0.0,
|
| 1259 |
+
"clip_ratio/region_mean": 0.0,
|
| 1260 |
+
"completions/clipped_ratio": 0.0,
|
| 1261 |
+
"completions/max_length": 117.2,
|
| 1262 |
+
"completions/max_terminated_length": 117.2,
|
| 1263 |
+
"completions/mean_length": 50.25,
|
| 1264 |
+
"completions/mean_terminated_length": 50.25,
|
| 1265 |
+
"completions/min_length": 15.4,
|
| 1266 |
+
"completions/min_terminated_length": 15.4,
|
| 1267 |
+
"entropy": 0.2198769545648247,
|
| 1268 |
+
"epoch": 0.010129310344827586,
|
| 1269 |
+
"frac_reward_zero_std": 0.1,
|
| 1270 |
+
"grad_norm": 0.5234375,
|
| 1271 |
+
"learning_rate": 3.1551724137931037e-06,
|
| 1272 |
+
"loss": -0.013682875037193298,
|
| 1273 |
+
"num_tokens": 551073.0,
|
| 1274 |
+
"reward": 0.4487079918384552,
|
| 1275 |
+
"reward_std": 0.5221681177150458,
|
| 1276 |
+
"rewards/reward_fn/mean": 0.4487079918384552,
|
| 1277 |
+
"rewards/reward_fn/std": 0.5221681545488537,
|
| 1278 |
+
"step": 235,
|
| 1279 |
+
"step_time": 20.431722840200383
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"clip_ratio/high_max": 0.0,
|
| 1283 |
+
"clip_ratio/high_mean": 0.0,
|
| 1284 |
+
"clip_ratio/low_mean": 0.0,
|
| 1285 |
+
"clip_ratio/low_min": 0.0,
|
| 1286 |
+
"clip_ratio/region_mean": 0.0,
|
| 1287 |
+
"completions/clipped_ratio": 0.0,
|
| 1288 |
+
"completions/max_length": 118.6,
|
| 1289 |
+
"completions/max_terminated_length": 118.6,
|
| 1290 |
+
"completions/mean_length": 54.675,
|
| 1291 |
+
"completions/mean_terminated_length": 54.675,
|
| 1292 |
+
"completions/min_length": 16.2,
|
| 1293 |
+
"completions/min_terminated_length": 16.2,
|
| 1294 |
+
"entropy": 0.20829223558394006,
|
| 1295 |
+
"epoch": 0.010344827586206896,
|
| 1296 |
+
"frac_reward_zero_std": 0.4,
|
| 1297 |
+
"grad_norm": 0.41015625,
|
| 1298 |
+
"learning_rate": 3.1120689655172413e-06,
|
| 1299 |
+
"loss": -0.030032938718795775,
|
| 1300 |
+
"num_tokens": 562420.0,
|
| 1301 |
+
"reward": 0.17456801533699035,
|
| 1302 |
+
"reward_std": 0.46873694062232973,
|
| 1303 |
+
"rewards/reward_fn/mean": 0.17456801533699035,
|
| 1304 |
+
"rewards/reward_fn/std": 0.4687369393184781,
|
| 1305 |
+
"step": 240,
|
| 1306 |
+
"step_time": 20.53446474100001
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"clip_ratio/high_max": 0.0,
|
| 1310 |
+
"clip_ratio/high_mean": 0.0,
|
| 1311 |
+
"clip_ratio/low_mean": 0.0,
|
| 1312 |
+
"clip_ratio/low_min": 0.0,
|
| 1313 |
+
"clip_ratio/region_mean": 0.0,
|
| 1314 |
+
"completions/clipped_ratio": 0.0,
|
| 1315 |
+
"completions/max_length": 73.8,
|
| 1316 |
+
"completions/max_terminated_length": 73.8,
|
| 1317 |
+
"completions/mean_length": 42.475,
|
| 1318 |
+
"completions/mean_terminated_length": 42.475,
|
| 1319 |
+
"completions/min_length": 15.8,
|
| 1320 |
+
"completions/min_terminated_length": 15.8,
|
| 1321 |
+
"entropy": 0.22942446915694745,
|
| 1322 |
+
"epoch": 0.010560344827586207,
|
| 1323 |
+
"frac_reward_zero_std": 0.4,
|
| 1324 |
+
"grad_norm": 0.0,
|
| 1325 |
+
"learning_rate": 3.0689655172413797e-06,
|
| 1326 |
+
"loss": 0.043540936708450315,
|
| 1327 |
+
"num_tokens": 572919.0,
|
| 1328 |
+
"reward": 0.5922660022974015,
|
| 1329 |
+
"reward_std": 0.26113789578084834,
|
| 1330 |
+
"rewards/reward_fn/mean": 0.5922660022974015,
|
| 1331 |
+
"rewards/reward_fn/std": 0.2611379227571888,
|
| 1332 |
+
"step": 245,
|
| 1333 |
+
"step_time": 14.755489835400112
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"clip_ratio/high_max": 0.0,
|
| 1337 |
+
"clip_ratio/high_mean": 0.0,
|
| 1338 |
+
"clip_ratio/low_mean": 0.0,
|
| 1339 |
+
"clip_ratio/low_min": 0.0,
|
| 1340 |
+
"clip_ratio/region_mean": 0.0,
|
| 1341 |
+
"completions/clipped_ratio": 0.0,
|
| 1342 |
+
"completions/max_length": 57.4,
|
| 1343 |
+
"completions/max_terminated_length": 57.4,
|
| 1344 |
+
"completions/mean_length": 31.925,
|
| 1345 |
+
"completions/mean_terminated_length": 31.925,
|
| 1346 |
+
"completions/min_length": 16.0,
|
| 1347 |
+
"completions/min_terminated_length": 16.0,
|
| 1348 |
+
"entropy": 0.16531706769019366,
|
| 1349 |
+
"epoch": 0.010775862068965518,
|
| 1350 |
+
"frac_reward_zero_std": 0.4,
|
| 1351 |
+
"grad_norm": 0.7734375,
|
| 1352 |
+
"learning_rate": 3.0258620689655178e-06,
|
| 1353 |
+
"loss": 0.02320311963558197,
|
| 1354 |
+
"num_tokens": 583104.0,
|
| 1355 |
+
"reward": 0.7694700002670288,
|
| 1356 |
+
"reward_std": 0.03285302005242556,
|
| 1357 |
+
"rewards/reward_fn/mean": 0.7694700002670288,
|
| 1358 |
+
"rewards/reward_fn/std": 0.03285302061121911,
|
| 1359 |
+
"step": 250,
|
| 1360 |
+
"step_time": 12.29388190760037
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"clip_ratio/high_max": 0.0,
|
| 1364 |
+
"clip_ratio/high_mean": 0.0,
|
| 1365 |
+
"clip_ratio/low_mean": 0.0,
|
| 1366 |
+
"clip_ratio/low_min": 0.0,
|
| 1367 |
+
"clip_ratio/region_mean": 0.0,
|
| 1368 |
+
"completions/clipped_ratio": 0.0,
|
| 1369 |
+
"completions/max_length": 62.2,
|
| 1370 |
+
"completions/max_terminated_length": 62.2,
|
| 1371 |
+
"completions/mean_length": 37.025,
|
| 1372 |
+
"completions/mean_terminated_length": 37.025,
|
| 1373 |
+
"completions/min_length": 15.6,
|
| 1374 |
+
"completions/min_terminated_length": 15.6,
|
| 1375 |
+
"entropy": 0.17417828142351938,
|
| 1376 |
+
"epoch": 0.010991379310344827,
|
| 1377 |
+
"frac_reward_zero_std": 0.4,
|
| 1378 |
+
"grad_norm": 0.546875,
|
| 1379 |
+
"learning_rate": 2.9827586206896554e-06,
|
| 1380 |
+
"loss": 0.0023472100496292115,
|
| 1381 |
+
"num_tokens": 593081.0,
|
| 1382 |
+
"reward": 0.6150040000677108,
|
| 1383 |
+
"reward_std": 0.19029689192248042,
|
| 1384 |
+
"rewards/reward_fn/mean": 0.6150040000677108,
|
| 1385 |
+
"rewards/reward_fn/std": 0.19029689457966015,
|
| 1386 |
+
"step": 255,
|
| 1387 |
+
"step_time": 12.741135484799088
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"clip_ratio/high_max": 0.0,
|
| 1391 |
+
"clip_ratio/high_mean": 0.0,
|
| 1392 |
+
"clip_ratio/low_mean": 0.0,
|
| 1393 |
+
"clip_ratio/low_min": 0.0,
|
| 1394 |
+
"clip_ratio/region_mean": 0.0,
|
| 1395 |
+
"completions/clipped_ratio": 0.0,
|
| 1396 |
+
"completions/max_length": 129.4,
|
| 1397 |
+
"completions/max_terminated_length": 129.4,
|
| 1398 |
+
"completions/mean_length": 63.825,
|
| 1399 |
+
"completions/mean_terminated_length": 63.825,
|
| 1400 |
+
"completions/min_length": 23.0,
|
| 1401 |
+
"completions/min_terminated_length": 23.0,
|
| 1402 |
+
"entropy": 0.253861751453951,
|
| 1403 |
+
"epoch": 0.011206896551724138,
|
| 1404 |
+
"frac_reward_zero_std": 0.4,
|
| 1405 |
+
"grad_norm": 0.20703125,
|
| 1406 |
+
"learning_rate": 2.9396551724137934e-06,
|
| 1407 |
+
"loss": 0.019092470407485962,
|
| 1408 |
+
"num_tokens": 605022.0,
|
| 1409 |
+
"reward": 0.5227800011634827,
|
| 1410 |
+
"reward_std": 0.22449488371494225,
|
| 1411 |
+
"rewards/reward_fn/mean": 0.5227800011634827,
|
| 1412 |
+
"rewards/reward_fn/std": 0.22449489417485893,
|
| 1413 |
+
"step": 260,
|
| 1414 |
+
"step_time": 21.9141057962006
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"clip_ratio/high_max": 0.0,
|
| 1418 |
+
"clip_ratio/high_mean": 0.0,
|
| 1419 |
+
"clip_ratio/low_mean": 0.0,
|
| 1420 |
+
"clip_ratio/low_min": 0.0,
|
| 1421 |
+
"clip_ratio/region_mean": 0.0,
|
| 1422 |
+
"completions/clipped_ratio": 0.0,
|
| 1423 |
+
"completions/max_length": 92.0,
|
| 1424 |
+
"completions/max_terminated_length": 92.0,
|
| 1425 |
+
"completions/mean_length": 35.05,
|
| 1426 |
+
"completions/mean_terminated_length": 35.05,
|
| 1427 |
+
"completions/min_length": 15.8,
|
| 1428 |
+
"completions/min_terminated_length": 15.8,
|
| 1429 |
+
"entropy": 0.1651882082573138,
|
| 1430 |
+
"epoch": 0.011422413793103449,
|
| 1431 |
+
"frac_reward_zero_std": 0.4,
|
| 1432 |
+
"grad_norm": 0.71875,
|
| 1433 |
+
"learning_rate": 2.8965517241379314e-06,
|
| 1434 |
+
"loss": -0.019537296891212464,
|
| 1435 |
+
"num_tokens": 615204.0,
|
| 1436 |
+
"reward": 0.6061379969120025,
|
| 1437 |
+
"reward_std": 0.3120082815119531,
|
| 1438 |
+
"rewards/reward_fn/mean": 0.6061379969120025,
|
| 1439 |
+
"rewards/reward_fn/std": 0.31200829716108275,
|
| 1440 |
+
"step": 265,
|
| 1441 |
+
"step_time": 17.125141170400457
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"clip_ratio/high_max": 0.0,
|
| 1445 |
+
"clip_ratio/high_mean": 0.0,
|
| 1446 |
+
"clip_ratio/low_mean": 0.0,
|
| 1447 |
+
"clip_ratio/low_min": 0.0,
|
| 1448 |
+
"clip_ratio/region_mean": 0.0,
|
| 1449 |
+
"completions/clipped_ratio": 0.0,
|
| 1450 |
+
"completions/max_length": 129.0,
|
| 1451 |
+
"completions/max_terminated_length": 129.0,
|
| 1452 |
+
"completions/mean_length": 56.975,
|
| 1453 |
+
"completions/mean_terminated_length": 56.975,
|
| 1454 |
+
"completions/min_length": 16.2,
|
| 1455 |
+
"completions/min_terminated_length": 16.2,
|
| 1456 |
+
"entropy": 0.24561482777353377,
|
| 1457 |
+
"epoch": 0.01163793103448276,
|
| 1458 |
+
"frac_reward_zero_std": 0.2,
|
| 1459 |
+
"grad_norm": 1.7421875,
|
| 1460 |
+
"learning_rate": 2.853448275862069e-06,
|
| 1461 |
+
"loss": 0.04291326403617859,
|
| 1462 |
+
"num_tokens": 626315.0,
|
| 1463 |
+
"reward": 0.18831600546836852,
|
| 1464 |
+
"reward_std": 0.5584077455103398,
|
| 1465 |
+
"rewards/reward_fn/mean": 0.18831600546836852,
|
| 1466 |
+
"rewards/reward_fn/std": 0.5584077499806881,
|
| 1467 |
+
"step": 270,
|
| 1468 |
+
"step_time": 21.739286486199852
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"clip_ratio/high_max": 0.0,
|
| 1472 |
+
"clip_ratio/high_mean": 0.0,
|
| 1473 |
+
"clip_ratio/low_mean": 0.0,
|
| 1474 |
+
"clip_ratio/low_min": 0.0,
|
| 1475 |
+
"clip_ratio/region_mean": 0.0,
|
| 1476 |
+
"completions/clipped_ratio": 0.0,
|
| 1477 |
+
"completions/max_length": 84.2,
|
| 1478 |
+
"completions/max_terminated_length": 84.2,
|
| 1479 |
+
"completions/mean_length": 57.575,
|
| 1480 |
+
"completions/mean_terminated_length": 57.575,
|
| 1481 |
+
"completions/min_length": 26.8,
|
| 1482 |
+
"completions/min_terminated_length": 26.8,
|
| 1483 |
+
"entropy": 0.23427433941978962,
|
| 1484 |
+
"epoch": 0.011853448275862068,
|
| 1485 |
+
"frac_reward_zero_std": 0.1,
|
| 1486 |
+
"grad_norm": 0.453125,
|
| 1487 |
+
"learning_rate": 2.810344827586207e-06,
|
| 1488 |
+
"loss": 0.039196133613586426,
|
| 1489 |
+
"num_tokens": 638030.0,
|
| 1490 |
+
"reward": 0.3995940089225769,
|
| 1491 |
+
"reward_std": 0.3814970766892657,
|
| 1492 |
+
"rewards/reward_fn/mean": 0.3995940089225769,
|
| 1493 |
+
"rewards/reward_fn/std": 0.3814970819279552,
|
| 1494 |
+
"step": 275,
|
| 1495 |
+
"step_time": 16.096988868400878
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"clip_ratio/high_max": 0.0,
|
| 1499 |
+
"clip_ratio/high_mean": 0.0,
|
| 1500 |
+
"clip_ratio/low_mean": 0.0,
|
| 1501 |
+
"clip_ratio/low_min": 0.0,
|
| 1502 |
+
"clip_ratio/region_mean": 0.0,
|
| 1503 |
+
"completions/clipped_ratio": 0.0,
|
| 1504 |
+
"completions/max_length": 168.0,
|
| 1505 |
+
"completions/max_terminated_length": 168.0,
|
| 1506 |
+
"completions/mean_length": 85.6,
|
| 1507 |
+
"completions/mean_terminated_length": 85.6,
|
| 1508 |
+
"completions/min_length": 34.0,
|
| 1509 |
+
"completions/min_terminated_length": 34.0,
|
| 1510 |
+
"entropy": 0.2679269164043944,
|
| 1511 |
+
"epoch": 0.01206896551724138,
|
| 1512 |
+
"frac_reward_zero_std": 0.4,
|
| 1513 |
+
"grad_norm": 0.0,
|
| 1514 |
+
"learning_rate": 2.7672413793103455e-06,
|
| 1515 |
+
"loss": 0.03390491306781769,
|
| 1516 |
+
"num_tokens": 650622.0,
|
| 1517 |
+
"reward": 0.2944400191307068,
|
| 1518 |
+
"reward_std": 0.6373479247093201,
|
| 1519 |
+
"rewards/reward_fn/mean": 0.2944400191307068,
|
| 1520 |
+
"rewards/reward_fn/std": 0.6373479485511779,
|
| 1521 |
+
"step": 280,
|
| 1522 |
+
"step_time": 26.78552907839967
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"clip_ratio/high_max": 0.0,
|
| 1526 |
+
"clip_ratio/high_mean": 0.0,
|
| 1527 |
+
"clip_ratio/low_mean": 0.0,
|
| 1528 |
+
"clip_ratio/low_min": 0.0,
|
| 1529 |
+
"clip_ratio/region_mean": 0.0,
|
| 1530 |
+
"completions/clipped_ratio": 0.025,
|
| 1531 |
+
"completions/max_length": 118.8,
|
| 1532 |
+
"completions/max_terminated_length": 111.2,
|
| 1533 |
+
"completions/mean_length": 59.8,
|
| 1534 |
+
"completions/mean_terminated_length": 55.75000152587891,
|
| 1535 |
+
"completions/min_length": 16.6,
|
| 1536 |
+
"completions/min_terminated_length": 16.6,
|
| 1537 |
+
"entropy": 0.2526095831533894,
|
| 1538 |
+
"epoch": 0.01228448275862069,
|
| 1539 |
+
"frac_reward_zero_std": 0.2,
|
| 1540 |
+
"grad_norm": 0.435546875,
|
| 1541 |
+
"learning_rate": 2.724137931034483e-06,
|
| 1542 |
+
"loss": 0.05690905451774597,
|
| 1543 |
+
"num_tokens": 662374.0,
|
| 1544 |
+
"reward": 0.1196260005235672,
|
| 1545 |
+
"reward_std": 0.767779788421467,
|
| 1546 |
+
"rewards/reward_fn/mean": 0.1196260005235672,
|
| 1547 |
+
"rewards/reward_fn/std": 0.7677798029500991,
|
| 1548 |
+
"step": 285,
|
| 1549 |
+
"step_time": 20.659876349599834
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"clip_ratio/high_max": 0.0,
|
| 1553 |
+
"clip_ratio/high_mean": 0.0,
|
| 1554 |
+
"clip_ratio/low_mean": 0.0,
|
| 1555 |
+
"clip_ratio/low_min": 0.0,
|
| 1556 |
+
"clip_ratio/region_mean": 0.0,
|
| 1557 |
+
"completions/clipped_ratio": 0.05,
|
| 1558 |
+
"completions/max_length": 122.0,
|
| 1559 |
+
"completions/max_terminated_length": 120.8,
|
| 1560 |
+
"completions/mean_length": 67.7,
|
| 1561 |
+
"completions/mean_terminated_length": 60.75833435058594,
|
| 1562 |
+
"completions/min_length": 15.4,
|
| 1563 |
+
"completions/min_terminated_length": 15.4,
|
| 1564 |
+
"entropy": 0.2572842217050493,
|
| 1565 |
+
"epoch": 0.0125,
|
| 1566 |
+
"frac_reward_zero_std": 0.2,
|
| 1567 |
+
"grad_norm": 0.302734375,
|
| 1568 |
+
"learning_rate": 2.681034482758621e-06,
|
| 1569 |
+
"loss": 0.08332639336585998,
|
| 1570 |
+
"num_tokens": 674146.0,
|
| 1571 |
+
"reward": -0.005963999032974243,
|
| 1572 |
+
"reward_std": 0.5873876449419185,
|
| 1573 |
+
"rewards/reward_fn/mean": -0.005963999032974243,
|
| 1574 |
+
"rewards/reward_fn/std": 0.5873876665486023,
|
| 1575 |
+
"step": 290,
|
| 1576 |
+
"step_time": 20.897897071199804
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"clip_ratio/high_max": 0.0,
|
| 1580 |
+
"clip_ratio/high_mean": 0.0,
|
| 1581 |
+
"clip_ratio/low_mean": 0.0,
|
| 1582 |
+
"clip_ratio/low_min": 0.0,
|
| 1583 |
+
"clip_ratio/region_mean": 0.0,
|
| 1584 |
+
"completions/clipped_ratio": 0.0,
|
| 1585 |
+
"completions/max_length": 83.0,
|
| 1586 |
+
"completions/max_terminated_length": 83.0,
|
| 1587 |
+
"completions/mean_length": 49.4,
|
| 1588 |
+
"completions/mean_terminated_length": 49.4,
|
| 1589 |
+
"completions/min_length": 24.2,
|
| 1590 |
+
"completions/min_terminated_length": 24.2,
|
| 1591 |
+
"entropy": 0.1755124410497956,
|
| 1592 |
+
"epoch": 0.01271551724137931,
|
| 1593 |
+
"frac_reward_zero_std": 0.4,
|
| 1594 |
+
"grad_norm": 0.63671875,
|
| 1595 |
+
"learning_rate": 2.637931034482759e-06,
|
| 1596 |
+
"loss": 0.02972428798675537,
|
| 1597 |
+
"num_tokens": 685162.0,
|
| 1598 |
+
"reward": 0.5233660042285919,
|
| 1599 |
+
"reward_std": 0.3009416783461347,
|
| 1600 |
+
"rewards/reward_fn/mean": 0.5233660042285919,
|
| 1601 |
+
"rewards/reward_fn/std": 0.3009416822576895,
|
| 1602 |
+
"step": 295,
|
| 1603 |
+
"step_time": 15.749949136399664
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"clip_ratio/high_max": 0.0,
|
| 1607 |
+
"clip_ratio/high_mean": 0.0,
|
| 1608 |
+
"clip_ratio/low_mean": 0.0,
|
| 1609 |
+
"clip_ratio/low_min": 0.0,
|
| 1610 |
+
"clip_ratio/region_mean": 0.0,
|
| 1611 |
+
"completions/clipped_ratio": 0.0,
|
| 1612 |
+
"completions/max_length": 93.8,
|
| 1613 |
+
"completions/max_terminated_length": 93.8,
|
| 1614 |
+
"completions/mean_length": 43.125,
|
| 1615 |
+
"completions/mean_terminated_length": 43.125,
|
| 1616 |
+
"completions/min_length": 16.0,
|
| 1617 |
+
"completions/min_terminated_length": 16.0,
|
| 1618 |
+
"entropy": 0.14239629595540465,
|
| 1619 |
+
"epoch": 0.01293103448275862,
|
| 1620 |
+
"frac_reward_zero_std": 0.6,
|
| 1621 |
+
"grad_norm": 0.29296875,
|
| 1622 |
+
"learning_rate": 2.5948275862068967e-06,
|
| 1623 |
+
"loss": 0.03822866082191467,
|
| 1624 |
+
"num_tokens": 696263.0,
|
| 1625 |
+
"reward": 0.7467899918556213,
|
| 1626 |
+
"reward_std": 0.14996536182879935,
|
| 1627 |
+
"rewards/reward_fn/mean": 0.7467899918556213,
|
| 1628 |
+
"rewards/reward_fn/std": 0.14996537954139058,
|
| 1629 |
+
"step": 300,
|
| 1630 |
+
"step_time": 17.312333570999545
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"clip_ratio/high_max": 0.0,
|
| 1634 |
+
"clip_ratio/high_mean": 0.0,
|
| 1635 |
+
"clip_ratio/low_mean": 0.0,
|
| 1636 |
+
"clip_ratio/low_min": 0.0,
|
| 1637 |
+
"clip_ratio/region_mean": 0.0,
|
| 1638 |
+
"completions/clipped_ratio": 0.025,
|
| 1639 |
+
"completions/max_length": 131.2,
|
| 1640 |
+
"completions/max_terminated_length": 100.6,
|
| 1641 |
+
"completions/mean_length": 55.55,
|
| 1642 |
+
"completions/mean_terminated_length": 50.860714721679685,
|
| 1643 |
+
"completions/min_length": 16.2,
|
| 1644 |
+
"completions/min_terminated_length": 16.2,
|
| 1645 |
+
"entropy": 0.22959837107919157,
|
| 1646 |
+
"epoch": 0.013146551724137931,
|
| 1647 |
+
"frac_reward_zero_std": 0.3,
|
| 1648 |
+
"grad_norm": 0.47265625,
|
| 1649 |
+
"learning_rate": 2.5517241379310347e-06,
|
| 1650 |
+
"loss": -0.094166100025177,
|
| 1651 |
+
"num_tokens": 707373.0,
|
| 1652 |
+
"reward": 0.13532000482082368,
|
| 1653 |
+
"reward_std": 0.8043574929237366,
|
| 1654 |
+
"rewards/reward_fn/mean": 0.13532000482082368,
|
| 1655 |
+
"rewards/reward_fn/std": 0.8043575048446655,
|
| 1656 |
+
"step": 305,
|
| 1657 |
+
"step_time": 21.792155743800322
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"clip_ratio/high_max": 0.0,
|
| 1661 |
+
"clip_ratio/high_mean": 0.0,
|
| 1662 |
+
"clip_ratio/low_mean": 0.0,
|
| 1663 |
+
"clip_ratio/low_min": 0.0,
|
| 1664 |
+
"clip_ratio/region_mean": 0.0,
|
| 1665 |
+
"completions/clipped_ratio": 0.025,
|
| 1666 |
+
"completions/max_length": 108.2,
|
| 1667 |
+
"completions/max_terminated_length": 89.8,
|
| 1668 |
+
"completions/mean_length": 53.7,
|
| 1669 |
+
"completions/mean_terminated_length": 49.05357208251953,
|
| 1670 |
+
"completions/min_length": 16.8,
|
| 1671 |
+
"completions/min_terminated_length": 16.8,
|
| 1672 |
+
"entropy": 0.17698723110370337,
|
| 1673 |
+
"epoch": 0.013362068965517242,
|
| 1674 |
+
"frac_reward_zero_std": 0.3,
|
| 1675 |
+
"grad_norm": 0.455078125,
|
| 1676 |
+
"learning_rate": 2.5086206896551723e-06,
|
| 1677 |
+
"loss": 0.054443192481994626,
|
| 1678 |
+
"num_tokens": 718741.0,
|
| 1679 |
+
"reward": 0.29295199513435366,
|
| 1680 |
+
"reward_std": 0.5932383094579563,
|
| 1681 |
+
"rewards/reward_fn/mean": 0.29295199513435366,
|
| 1682 |
+
"rewards/reward_fn/std": 0.5932383131046663,
|
| 1683 |
+
"step": 310,
|
| 1684 |
+
"step_time": 19.077739837599985
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"clip_ratio/high_max": 0.0,
|
| 1688 |
+
"clip_ratio/high_mean": 0.0,
|
| 1689 |
+
"clip_ratio/low_mean": 0.0,
|
| 1690 |
+
"clip_ratio/low_min": 0.0,
|
| 1691 |
+
"clip_ratio/region_mean": 0.0,
|
| 1692 |
+
"completions/clipped_ratio": 0.15,
|
| 1693 |
+
"completions/max_length": 159.6,
|
| 1694 |
+
"completions/max_terminated_length": 136.0,
|
| 1695 |
+
"completions/mean_length": 92.8,
|
| 1696 |
+
"completions/mean_terminated_length": 73.96666717529297,
|
| 1697 |
+
"completions/min_length": 34.8,
|
| 1698 |
+
"completions/min_terminated_length": 34.8,
|
| 1699 |
+
"entropy": 0.1539902521762997,
|
| 1700 |
+
"epoch": 0.013577586206896551,
|
| 1701 |
+
"frac_reward_zero_std": 0.4,
|
| 1702 |
+
"grad_norm": 0.1796875,
|
| 1703 |
+
"learning_rate": 2.4655172413793104e-06,
|
| 1704 |
+
"loss": 0.04268620014190674,
|
| 1705 |
+
"num_tokens": 732317.0,
|
| 1706 |
+
"reward": 0.09846000373363495,
|
| 1707 |
+
"reward_std": 0.654845405537344,
|
| 1708 |
+
"rewards/reward_fn/mean": 0.09846000373363495,
|
| 1709 |
+
"rewards/reward_fn/std": 0.6548454073446919,
|
| 1710 |
+
"step": 315,
|
| 1711 |
+
"step_time": 25.857181660000606
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"clip_ratio/high_max": 0.0,
|
| 1715 |
+
"clip_ratio/high_mean": 0.0,
|
| 1716 |
+
"clip_ratio/low_mean": 0.0,
|
| 1717 |
+
"clip_ratio/low_min": 0.0,
|
| 1718 |
+
"clip_ratio/region_mean": 0.0,
|
| 1719 |
+
"completions/clipped_ratio": 0.0,
|
| 1720 |
+
"completions/max_length": 46.8,
|
| 1721 |
+
"completions/max_terminated_length": 46.8,
|
| 1722 |
+
"completions/mean_length": 29.525,
|
| 1723 |
+
"completions/mean_terminated_length": 29.525,
|
| 1724 |
+
"completions/min_length": 16.0,
|
| 1725 |
+
"completions/min_terminated_length": 16.0,
|
| 1726 |
+
"entropy": 0.13891502455808222,
|
| 1727 |
+
"epoch": 0.013793103448275862,
|
| 1728 |
+
"frac_reward_zero_std": 0.5,
|
| 1729 |
+
"grad_norm": 0.0,
|
| 1730 |
+
"learning_rate": 2.4224137931034484e-06,
|
| 1731 |
+
"loss": 0.0070524528622627255,
|
| 1732 |
+
"num_tokens": 742546.0,
|
| 1733 |
+
"reward": 0.5995699942111969,
|
| 1734 |
+
"reward_std": 0.21884493981051492,
|
| 1735 |
+
"rewards/reward_fn/mean": 0.5995699942111969,
|
| 1736 |
+
"rewards/reward_fn/std": 0.21884493713005213,
|
| 1737 |
+
"step": 320,
|
| 1738 |
+
"step_time": 11.21470864460007
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"clip_ratio/high_max": 0.0,
|
| 1742 |
+
"clip_ratio/high_mean": 0.0,
|
| 1743 |
+
"clip_ratio/low_mean": 0.0,
|
| 1744 |
+
"clip_ratio/low_min": 0.0,
|
| 1745 |
+
"clip_ratio/region_mean": 0.0,
|
| 1746 |
+
"completions/clipped_ratio": 0.075,
|
| 1747 |
+
"completions/max_length": 107.2,
|
| 1748 |
+
"completions/max_terminated_length": 105.4,
|
| 1749 |
+
"completions/mean_length": 58.825,
|
| 1750 |
+
"completions/mean_terminated_length": 45.85,
|
| 1751 |
+
"completions/min_length": 16.2,
|
| 1752 |
+
"completions/min_terminated_length": 16.2,
|
| 1753 |
+
"entropy": 0.18881899882107972,
|
| 1754 |
+
"epoch": 0.014008620689655173,
|
| 1755 |
+
"frac_reward_zero_std": 0.4,
|
| 1756 |
+
"grad_norm": 1.0234375,
|
| 1757 |
+
"learning_rate": 2.3793103448275864e-06,
|
| 1758 |
+
"loss": -0.006822568923234939,
|
| 1759 |
+
"num_tokens": 754463.0,
|
| 1760 |
+
"reward": 0.05360400676727295,
|
| 1761 |
+
"reward_std": 0.8347343930785428,
|
| 1762 |
+
"rewards/reward_fn/mean": 0.05360400676727295,
|
| 1763 |
+
"rewards/reward_fn/std": 0.8347344425696065,
|
| 1764 |
+
"step": 325,
|
| 1765 |
+
"step_time": 19.000760145999447
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"clip_ratio/high_max": 0.0,
|
| 1769 |
+
"clip_ratio/high_mean": 0.0,
|
| 1770 |
+
"clip_ratio/low_mean": 0.0,
|
| 1771 |
+
"clip_ratio/low_min": 0.0,
|
| 1772 |
+
"clip_ratio/region_mean": 0.0,
|
| 1773 |
+
"completions/clipped_ratio": 0.05,
|
| 1774 |
+
"completions/max_length": 137.8,
|
| 1775 |
+
"completions/max_terminated_length": 133.2,
|
| 1776 |
+
"completions/mean_length": 76.525,
|
| 1777 |
+
"completions/mean_terminated_length": 70.575,
|
| 1778 |
+
"completions/min_length": 29.4,
|
| 1779 |
+
"completions/min_terminated_length": 29.4,
|
| 1780 |
+
"entropy": 0.2549753186496673,
|
| 1781 |
+
"epoch": 0.014224137931034483,
|
| 1782 |
+
"frac_reward_zero_std": 0.5,
|
| 1783 |
+
"grad_norm": 0.291015625,
|
| 1784 |
+
"learning_rate": 2.3362068965517244e-06,
|
| 1785 |
+
"loss": -0.00032879412174224854,
|
| 1786 |
+
"num_tokens": 766976.0,
|
| 1787 |
+
"reward": 0.15794800221920013,
|
| 1788 |
+
"reward_std": 0.39052192161325366,
|
| 1789 |
+
"rewards/reward_fn/mean": 0.15794800221920013,
|
| 1790 |
+
"rewards/reward_fn/std": 0.39052194436080756,
|
| 1791 |
+
"step": 330,
|
| 1792 |
+
"step_time": 23.229899759599355
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"clip_ratio/high_max": 0.0,
|
| 1796 |
+
"clip_ratio/high_mean": 0.0,
|
| 1797 |
+
"clip_ratio/low_mean": 0.0,
|
| 1798 |
+
"clip_ratio/low_min": 0.0,
|
| 1799 |
+
"clip_ratio/region_mean": 0.0,
|
| 1800 |
+
"completions/clipped_ratio": 0.0,
|
| 1801 |
+
"completions/max_length": 113.8,
|
| 1802 |
+
"completions/max_terminated_length": 113.8,
|
| 1803 |
+
"completions/mean_length": 44.9,
|
| 1804 |
+
"completions/mean_terminated_length": 44.9,
|
| 1805 |
+
"completions/min_length": 16.8,
|
| 1806 |
+
"completions/min_terminated_length": 16.8,
|
| 1807 |
+
"entropy": 0.16029987429210452,
|
| 1808 |
+
"epoch": 0.014439655172413792,
|
| 1809 |
+
"frac_reward_zero_std": 0.5,
|
| 1810 |
+
"grad_norm": 0.0,
|
| 1811 |
+
"learning_rate": 2.293103448275862e-06,
|
| 1812 |
+
"loss": -0.1327621340751648,
|
| 1813 |
+
"num_tokens": 777720.0,
|
| 1814 |
+
"reward": -0.13512398451566696,
|
| 1815 |
+
"reward_std": 0.6157536920160055,
|
| 1816 |
+
"rewards/reward_fn/mean": -0.13512398451566696,
|
| 1817 |
+
"rewards/reward_fn/std": 0.6157537158578634,
|
| 1818 |
+
"step": 335,
|
| 1819 |
+
"step_time": 19.72434388899965
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"clip_ratio/high_max": 0.0,
|
| 1823 |
+
"clip_ratio/high_mean": 0.0,
|
| 1824 |
+
"clip_ratio/low_mean": 0.0,
|
| 1825 |
+
"clip_ratio/low_min": 0.0,
|
| 1826 |
+
"clip_ratio/region_mean": 0.0,
|
| 1827 |
+
"completions/clipped_ratio": 0.15,
|
| 1828 |
+
"completions/max_length": 124.8,
|
| 1829 |
+
"completions/max_terminated_length": 108.0,
|
| 1830 |
+
"completions/mean_length": 89.45,
|
| 1831 |
+
"completions/mean_terminated_length": 73.82000122070312,
|
| 1832 |
+
"completions/min_length": 48.4,
|
| 1833 |
+
"completions/min_terminated_length": 48.4,
|
| 1834 |
+
"entropy": 0.22397661422437523,
|
| 1835 |
+
"epoch": 0.014655172413793103,
|
| 1836 |
+
"frac_reward_zero_std": 0.4,
|
| 1837 |
+
"grad_norm": 0.234375,
|
| 1838 |
+
"learning_rate": 2.25e-06,
|
| 1839 |
+
"loss": 0.020370352268218993,
|
| 1840 |
+
"num_tokens": 791050.0,
|
| 1841 |
+
"reward": 0.02506202459335327,
|
| 1842 |
+
"reward_std": 0.030776232342759613,
|
| 1843 |
+
"rewards/reward_fn/mean": 0.02506202459335327,
|
| 1844 |
+
"rewards/reward_fn/std": 0.03077622954297112,
|
| 1845 |
+
"step": 340,
|
| 1846 |
+
"step_time": 21.360666513999785
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"clip_ratio/high_max": 0.0,
|
| 1850 |
+
"clip_ratio/high_mean": 0.0,
|
| 1851 |
+
"clip_ratio/low_mean": 0.0,
|
| 1852 |
+
"clip_ratio/low_min": 0.0,
|
| 1853 |
+
"clip_ratio/region_mean": 0.0,
|
| 1854 |
+
"completions/clipped_ratio": 0.125,
|
| 1855 |
+
"completions/max_length": 163.8,
|
| 1856 |
+
"completions/max_terminated_length": 127.6,
|
| 1857 |
+
"completions/mean_length": 75.875,
|
| 1858 |
+
"completions/mean_terminated_length": 50.14833374023438,
|
| 1859 |
+
"completions/min_length": 15.8,
|
| 1860 |
+
"completions/min_terminated_length": 15.8,
|
| 1861 |
+
"entropy": 0.14475313751026989,
|
| 1862 |
+
"epoch": 0.014870689655172414,
|
| 1863 |
+
"frac_reward_zero_std": 0.4,
|
| 1864 |
+
"grad_norm": 0.490234375,
|
| 1865 |
+
"learning_rate": 2.206896551724138e-06,
|
| 1866 |
+
"loss": 0.08303472399711609,
|
| 1867 |
+
"num_tokens": 803649.0,
|
| 1868 |
+
"reward": 0.38839800357818605,
|
| 1869 |
+
"reward_std": 0.44462206639145735,
|
| 1870 |
+
"rewards/reward_fn/mean": 0.38839800357818605,
|
| 1871 |
+
"rewards/reward_fn/std": 0.44462209131597774,
|
| 1872 |
+
"step": 345,
|
| 1873 |
+
"step_time": 26.33779034660056
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"clip_ratio/high_max": 0.0,
|
| 1877 |
+
"clip_ratio/high_mean": 0.0,
|
| 1878 |
+
"clip_ratio/low_mean": 0.0,
|
| 1879 |
+
"clip_ratio/low_min": 0.0,
|
| 1880 |
+
"clip_ratio/region_mean": 0.0,
|
| 1881 |
+
"completions/clipped_ratio": 0.1,
|
| 1882 |
+
"completions/max_length": 186.8,
|
| 1883 |
+
"completions/max_terminated_length": 176.0,
|
| 1884 |
+
"completions/mean_length": 85.5,
|
| 1885 |
+
"completions/mean_terminated_length": 67.71666870117187,
|
| 1886 |
+
"completions/min_length": 16.0,
|
| 1887 |
+
"completions/min_terminated_length": 16.0,
|
| 1888 |
+
"entropy": 0.233237065654248,
|
| 1889 |
+
"epoch": 0.015086206896551725,
|
| 1890 |
+
"frac_reward_zero_std": 0.3,
|
| 1891 |
+
"grad_norm": 0.07177734375,
|
| 1892 |
+
"learning_rate": 2.163793103448276e-06,
|
| 1893 |
+
"loss": -0.04608747363090515,
|
| 1894 |
+
"num_tokens": 816513.0,
|
| 1895 |
+
"reward": -0.31583799719810485,
|
| 1896 |
+
"reward_std": 0.6374753089621663,
|
| 1897 |
+
"rewards/reward_fn/mean": -0.31583799719810485,
|
| 1898 |
+
"rewards/reward_fn/std": 0.637475342862308,
|
| 1899 |
+
"step": 350,
|
| 1900 |
+
"step_time": 29.608003718000145
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"clip_ratio/high_max": 0.0,
|
| 1904 |
+
"clip_ratio/high_mean": 0.0,
|
| 1905 |
+
"clip_ratio/low_mean": 0.0,
|
| 1906 |
+
"clip_ratio/low_min": 0.0,
|
| 1907 |
+
"clip_ratio/region_mean": 0.0,
|
| 1908 |
+
"completions/clipped_ratio": 0.05,
|
| 1909 |
+
"completions/max_length": 143.4,
|
| 1910 |
+
"completions/max_terminated_length": 110.8,
|
| 1911 |
+
"completions/mean_length": 61.4,
|
| 1912 |
+
"completions/mean_terminated_length": 52.72500152587891,
|
| 1913 |
+
"completions/min_length": 16.4,
|
| 1914 |
+
"completions/min_terminated_length": 16.4,
|
| 1915 |
+
"entropy": 0.20117253810167313,
|
| 1916 |
+
"epoch": 0.015301724137931034,
|
| 1917 |
+
"frac_reward_zero_std": 0.4,
|
| 1918 |
+
"grad_norm": 0.0,
|
| 1919 |
+
"learning_rate": 2.120689655172414e-06,
|
| 1920 |
+
"loss": -0.0012839555740356446,
|
| 1921 |
+
"num_tokens": 828305.0,
|
| 1922 |
+
"reward": 0.11507600247859955,
|
| 1923 |
+
"reward_std": 0.5993056967010488,
|
| 1924 |
+
"rewards/reward_fn/mean": 0.11507600247859955,
|
| 1925 |
+
"rewards/reward_fn/std": 0.5993057217798196,
|
| 1926 |
+
"step": 355,
|
| 1927 |
+
"step_time": 23.633322461799253
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"clip_ratio/high_max": 0.0,
|
| 1931 |
+
"clip_ratio/high_mean": 0.0,
|
| 1932 |
+
"clip_ratio/low_mean": 0.0,
|
| 1933 |
+
"clip_ratio/low_min": 0.0,
|
| 1934 |
+
"clip_ratio/region_mean": 0.0,
|
| 1935 |
+
"completions/clipped_ratio": 0.0,
|
| 1936 |
+
"completions/max_length": 63.4,
|
| 1937 |
+
"completions/max_terminated_length": 63.4,
|
| 1938 |
+
"completions/mean_length": 30.9,
|
| 1939 |
+
"completions/mean_terminated_length": 30.9,
|
| 1940 |
+
"completions/min_length": 15.6,
|
| 1941 |
+
"completions/min_terminated_length": 15.6,
|
| 1942 |
+
"entropy": 0.17519649318710434,
|
| 1943 |
+
"epoch": 0.015517241379310345,
|
| 1944 |
+
"frac_reward_zero_std": 0.4,
|
| 1945 |
+
"grad_norm": 0.99609375,
|
| 1946 |
+
"learning_rate": 2.0775862068965517e-06,
|
| 1947 |
+
"loss": 0.025579386949539186,
|
| 1948 |
+
"num_tokens": 838501.0,
|
| 1949 |
+
"reward": 0.4936339855194092,
|
| 1950 |
+
"reward_std": 0.39777664675784763,
|
| 1951 |
+
"rewards/reward_fn/mean": 0.4936339855194092,
|
| 1952 |
+
"rewards/reward_fn/std": 0.3977766572148539,
|
| 1953 |
+
"step": 360,
|
| 1954 |
+
"step_time": 13.110204230200543
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"clip_ratio/high_max": 0.0,
|
| 1958 |
+
"clip_ratio/high_mean": 0.0,
|
| 1959 |
+
"clip_ratio/low_mean": 0.0,
|
| 1960 |
+
"clip_ratio/low_min": 0.0,
|
| 1961 |
+
"clip_ratio/region_mean": 0.0,
|
| 1962 |
+
"completions/clipped_ratio": 0.0,
|
| 1963 |
+
"completions/max_length": 61.6,
|
| 1964 |
+
"completions/max_terminated_length": 61.6,
|
| 1965 |
+
"completions/mean_length": 30.575,
|
| 1966 |
+
"completions/mean_terminated_length": 30.575,
|
| 1967 |
+
"completions/min_length": 16.6,
|
| 1968 |
+
"completions/min_terminated_length": 16.6,
|
| 1969 |
+
"entropy": 0.14497978072613477,
|
| 1970 |
+
"epoch": 0.015732758620689655,
|
| 1971 |
+
"frac_reward_zero_std": 0.7,
|
| 1972 |
+
"grad_norm": 0.43359375,
|
| 1973 |
+
"learning_rate": 2.0344827586206897e-06,
|
| 1974 |
+
"loss": 0.05366594791412353,
|
| 1975 |
+
"num_tokens": 848536.0,
|
| 1976 |
+
"reward": 0.3110899955034256,
|
| 1977 |
+
"reward_std": 0.5116304856666829,
|
| 1978 |
+
"rewards/reward_fn/mean": 0.3110899955034256,
|
| 1979 |
+
"rewards/reward_fn/std": 0.5116304916300578,
|
| 1980 |
+
"step": 365,
|
| 1981 |
+
"step_time": 12.62919970000039
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"clip_ratio/high_max": 0.0,
|
| 1985 |
+
"clip_ratio/high_mean": 0.0,
|
| 1986 |
+
"clip_ratio/low_mean": 0.0,
|
| 1987 |
+
"clip_ratio/low_min": 0.0,
|
| 1988 |
+
"clip_ratio/region_mean": 0.0,
|
| 1989 |
+
"completions/clipped_ratio": 0.0,
|
| 1990 |
+
"completions/max_length": 82.6,
|
| 1991 |
+
"completions/max_terminated_length": 82.6,
|
| 1992 |
+
"completions/mean_length": 40.325,
|
| 1993 |
+
"completions/mean_terminated_length": 40.325,
|
| 1994 |
+
"completions/min_length": 15.8,
|
| 1995 |
+
"completions/min_terminated_length": 15.8,
|
| 1996 |
+
"entropy": 0.1809167622952373,
|
| 1997 |
+
"epoch": 0.015948275862068966,
|
| 1998 |
+
"frac_reward_zero_std": 0.4,
|
| 1999 |
+
"grad_norm": 1.765625,
|
| 2000 |
+
"learning_rate": 1.9913793103448278e-06,
|
| 2001 |
+
"loss": 0.04847588837146759,
|
| 2002 |
+
"num_tokens": 859473.0,
|
| 2003 |
+
"reward": 0.2660140126943588,
|
| 2004 |
+
"reward_std": 0.645512792468071,
|
| 2005 |
+
"rewards/reward_fn/mean": 0.2660140126943588,
|
| 2006 |
+
"rewards/reward_fn/std": 0.645512792468071,
|
| 2007 |
+
"step": 370,
|
| 2008 |
+
"step_time": 16.079390238599444
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"clip_ratio/high_max": 0.0,
|
| 2012 |
+
"clip_ratio/high_mean": 0.0,
|
| 2013 |
+
"clip_ratio/low_mean": 0.0,
|
| 2014 |
+
"clip_ratio/low_min": 0.0,
|
| 2015 |
+
"clip_ratio/region_mean": 0.0,
|
| 2016 |
+
"completions/clipped_ratio": 0.0,
|
| 2017 |
+
"completions/max_length": 52.0,
|
| 2018 |
+
"completions/max_terminated_length": 52.0,
|
| 2019 |
+
"completions/mean_length": 24.85,
|
| 2020 |
+
"completions/mean_terminated_length": 24.85,
|
| 2021 |
+
"completions/min_length": 15.8,
|
| 2022 |
+
"completions/min_terminated_length": 15.8,
|
| 2023 |
+
"entropy": 0.18615856370088296,
|
| 2024 |
+
"epoch": 0.016163793103448277,
|
| 2025 |
+
"frac_reward_zero_std": 0.5,
|
| 2026 |
+
"grad_norm": 0.4375,
|
| 2027 |
+
"learning_rate": 1.948275862068966e-06,
|
| 2028 |
+
"loss": -0.12370493412017822,
|
| 2029 |
+
"num_tokens": 868927.0,
|
| 2030 |
+
"reward": 0.3794360011816025,
|
| 2031 |
+
"reward_std": 0.5513797464169329,
|
| 2032 |
+
"rewards/reward_fn/mean": 0.3794360011816025,
|
| 2033 |
+
"rewards/reward_fn/std": 0.5513797624618746,
|
| 2034 |
+
"step": 375,
|
| 2035 |
+
"step_time": 11.392414143199858
|
| 2036 |
+
},
|
| 2037 |
+
{
|
| 2038 |
+
"clip_ratio/high_max": 0.0,
|
| 2039 |
+
"clip_ratio/high_mean": 0.0,
|
| 2040 |
+
"clip_ratio/low_mean": 0.0,
|
| 2041 |
+
"clip_ratio/low_min": 0.0,
|
| 2042 |
+
"clip_ratio/region_mean": 0.0,
|
| 2043 |
+
"completions/clipped_ratio": 0.0,
|
| 2044 |
+
"completions/max_length": 103.0,
|
| 2045 |
+
"completions/max_terminated_length": 103.0,
|
| 2046 |
+
"completions/mean_length": 62.25,
|
| 2047 |
+
"completions/mean_terminated_length": 62.25,
|
| 2048 |
+
"completions/min_length": 32.8,
|
| 2049 |
+
"completions/min_terminated_length": 32.8,
|
| 2050 |
+
"entropy": 0.12981002542364878,
|
| 2051 |
+
"epoch": 0.016379310344827588,
|
| 2052 |
+
"frac_reward_zero_std": 0.5,
|
| 2053 |
+
"grad_norm": 0.0,
|
| 2054 |
+
"learning_rate": 1.9051724137931036e-06,
|
| 2055 |
+
"loss": -0.005396611988544464,
|
| 2056 |
+
"num_tokens": 880641.0,
|
| 2057 |
+
"reward": 0.37234801054000854,
|
| 2058 |
+
"reward_std": 0.33126811534166334,
|
| 2059 |
+
"rewards/reward_fn/mean": 0.37234801054000854,
|
| 2060 |
+
"rewards/reward_fn/std": 0.33126811534166334,
|
| 2061 |
+
"step": 380,
|
| 2062 |
+
"step_time": 18.616042540200944
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"clip_ratio/high_max": 0.0,
|
| 2066 |
+
"clip_ratio/high_mean": 0.0,
|
| 2067 |
+
"clip_ratio/low_mean": 0.0,
|
| 2068 |
+
"clip_ratio/low_min": 0.0,
|
| 2069 |
+
"clip_ratio/region_mean": 0.0,
|
| 2070 |
+
"completions/clipped_ratio": 0.0,
|
| 2071 |
+
"completions/max_length": 54.2,
|
| 2072 |
+
"completions/max_terminated_length": 54.2,
|
| 2073 |
+
"completions/mean_length": 37.275,
|
| 2074 |
+
"completions/mean_terminated_length": 37.275,
|
| 2075 |
+
"completions/min_length": 16.6,
|
| 2076 |
+
"completions/min_terminated_length": 16.6,
|
| 2077 |
+
"entropy": 0.11609641392715275,
|
| 2078 |
+
"epoch": 0.016594827586206895,
|
| 2079 |
+
"frac_reward_zero_std": 0.4,
|
| 2080 |
+
"grad_norm": 0.67578125,
|
| 2081 |
+
"learning_rate": 1.8620689655172416e-06,
|
| 2082 |
+
"loss": 0.0155934676527977,
|
| 2083 |
+
"num_tokens": 891240.0,
|
| 2084 |
+
"reward": 0.5631299942731858,
|
| 2085 |
+
"reward_std": 0.35218784370808864,
|
| 2086 |
+
"rewards/reward_fn/mean": 0.5631299942731858,
|
| 2087 |
+
"rewards/reward_fn/std": 0.35218784669123127,
|
| 2088 |
+
"step": 385,
|
| 2089 |
+
"step_time": 11.740256453400798
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"clip_ratio/high_max": 0.0,
|
| 2093 |
+
"clip_ratio/high_mean": 0.0,
|
| 2094 |
+
"clip_ratio/low_mean": 0.0,
|
| 2095 |
+
"clip_ratio/low_min": 0.0,
|
| 2096 |
+
"clip_ratio/region_mean": 0.0,
|
| 2097 |
+
"completions/clipped_ratio": 0.0,
|
| 2098 |
+
"completions/max_length": 112.8,
|
| 2099 |
+
"completions/max_terminated_length": 112.8,
|
| 2100 |
+
"completions/mean_length": 51.875,
|
| 2101 |
+
"completions/mean_terminated_length": 51.875,
|
| 2102 |
+
"completions/min_length": 17.4,
|
| 2103 |
+
"completions/min_terminated_length": 17.4,
|
| 2104 |
+
"entropy": 0.13740509356721303,
|
| 2105 |
+
"epoch": 0.016810344827586206,
|
| 2106 |
+
"frac_reward_zero_std": 0.3,
|
| 2107 |
+
"grad_norm": 0.49609375,
|
| 2108 |
+
"learning_rate": 1.8189655172413794e-06,
|
| 2109 |
+
"loss": 0.10463507175445556,
|
| 2110 |
+
"num_tokens": 902783.0,
|
| 2111 |
+
"reward": 0.6312699973583221,
|
| 2112 |
+
"reward_std": 0.2348479055101052,
|
| 2113 |
+
"rewards/reward_fn/mean": 0.6312699973583221,
|
| 2114 |
+
"rewards/reward_fn/std": 0.23484793551033362,
|
| 2115 |
+
"step": 390,
|
| 2116 |
+
"step_time": 19.62515979279924
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"clip_ratio/high_max": 0.0,
|
| 2120 |
+
"clip_ratio/high_mean": 0.0,
|
| 2121 |
+
"clip_ratio/low_mean": 0.0,
|
| 2122 |
+
"clip_ratio/low_min": 0.0,
|
| 2123 |
+
"clip_ratio/region_mean": 0.0,
|
| 2124 |
+
"completions/clipped_ratio": 0.0,
|
| 2125 |
+
"completions/max_length": 68.8,
|
| 2126 |
+
"completions/max_terminated_length": 68.8,
|
| 2127 |
+
"completions/mean_length": 38.425,
|
| 2128 |
+
"completions/mean_terminated_length": 38.425,
|
| 2129 |
+
"completions/min_length": 16.2,
|
| 2130 |
+
"completions/min_terminated_length": 16.2,
|
| 2131 |
+
"entropy": 0.22300243815625437,
|
| 2132 |
+
"epoch": 0.017025862068965517,
|
| 2133 |
+
"frac_reward_zero_std": 0.5,
|
| 2134 |
+
"grad_norm": 0.376953125,
|
| 2135 |
+
"learning_rate": 1.7758620689655172e-06,
|
| 2136 |
+
"loss": 0.020081929862499237,
|
| 2137 |
+
"num_tokens": 913372.0,
|
| 2138 |
+
"reward": 0.3120719850063324,
|
| 2139 |
+
"reward_std": 0.6087781610287493,
|
| 2140 |
+
"rewards/reward_fn/mean": 0.3120719850063324,
|
| 2141 |
+
"rewards/reward_fn/std": 0.6087781770736911,
|
| 2142 |
+
"step": 395,
|
| 2143 |
+
"step_time": 13.894178912400093
|
| 2144 |
+
},
|
| 2145 |
+
{
|
| 2146 |
+
"clip_ratio/high_max": 0.0,
|
| 2147 |
+
"clip_ratio/high_mean": 0.0,
|
| 2148 |
+
"clip_ratio/low_mean": 0.0,
|
| 2149 |
+
"clip_ratio/low_min": 0.0,
|
| 2150 |
+
"clip_ratio/region_mean": 0.0,
|
| 2151 |
+
"completions/clipped_ratio": 0.0,
|
| 2152 |
+
"completions/max_length": 42.4,
|
| 2153 |
+
"completions/max_terminated_length": 42.4,
|
| 2154 |
+
"completions/mean_length": 26.6,
|
| 2155 |
+
"completions/mean_terminated_length": 26.6,
|
| 2156 |
+
"completions/min_length": 15.8,
|
| 2157 |
+
"completions/min_terminated_length": 15.8,
|
| 2158 |
+
"entropy": 0.16512603219016456,
|
| 2159 |
+
"epoch": 0.017241379310344827,
|
| 2160 |
+
"frac_reward_zero_std": 0.5,
|
| 2161 |
+
"grad_norm": 0.71484375,
|
| 2162 |
+
"learning_rate": 1.7327586206896555e-06,
|
| 2163 |
+
"loss": -0.00851401463150978,
|
| 2164 |
+
"num_tokens": 923000.0,
|
| 2165 |
+
"reward": 0.1584160089492798,
|
| 2166 |
+
"reward_std": 0.7910981756576803,
|
| 2167 |
+
"rewards/reward_fn/mean": 0.1584160089492798,
|
| 2168 |
+
"rewards/reward_fn/std": 0.7910982114233776,
|
| 2169 |
+
"step": 400,
|
| 2170 |
+
"step_time": 10.169051740999567
|
| 2171 |
+
}
|
| 2172 |
+
],
|
| 2173 |
+
"logging_steps": 5,
|
| 2174 |
+
"max_steps": 600,
|
| 2175 |
+
"num_input_tokens_seen": 923000,
|
| 2176 |
+
"num_train_epochs": 1,
|
| 2177 |
+
"save_steps": 50,
|
| 2178 |
+
"stateful_callbacks": {
|
| 2179 |
+
"TrainerControl": {
|
| 2180 |
+
"args": {
|
| 2181 |
+
"should_epoch_stop": false,
|
| 2182 |
+
"should_evaluate": false,
|
| 2183 |
+
"should_log": false,
|
| 2184 |
+
"should_save": true,
|
| 2185 |
+
"should_training_stop": false
|
| 2186 |
+
},
|
| 2187 |
+
"attributes": {}
|
| 2188 |
+
}
|
| 2189 |
+
},
|
| 2190 |
+
"total_flos": 0.0,
|
| 2191 |
+
"train_batch_size": 1,
|
| 2192 |
+
"trial_name": null,
|
| 2193 |
+
"trial_params": null
|
| 2194 |
+
}
|
checkpoint-400/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0cf50531ca359d39fbfa05dc0896e03532a83d53b029182b6d7f757efab0c97a
|
| 3 |
+
size 7185
|