Upload task output 8225782d-8d83-4e96-88c2-5e335b20e67c
Browse files- config.json +31 -0
- generation_config.json +7 -0
- loss.txt +1 -0
- model.safetensors +3 -0
- special_tokens_map.json +30 -0
- tokenizer.json +0 -0
- tokenizer.model +3 -0
- tokenizer_config.json +44 -0
- trainer_state.json +2183 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"LlamaForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 1,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"head_dim": 64,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 2048,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 5632,
|
| 14 |
+
"max_position_embeddings": 2048,
|
| 15 |
+
"mlp_bias": false,
|
| 16 |
+
"model_type": "llama",
|
| 17 |
+
"num_attention_heads": 32,
|
| 18 |
+
"num_hidden_layers": 22,
|
| 19 |
+
"num_key_value_heads": 4,
|
| 20 |
+
"pad_token_id": 0,
|
| 21 |
+
"pretraining_tp": 1,
|
| 22 |
+
"rms_norm_eps": 1e-05,
|
| 23 |
+
"rope_scaling": null,
|
| 24 |
+
"rope_theta": 10000.0,
|
| 25 |
+
"tie_word_embeddings": false,
|
| 26 |
+
"torch_dtype": "bfloat16",
|
| 27 |
+
"transformers_version": "4.51.3",
|
| 28 |
+
"unsloth_version": "2024.9",
|
| 29 |
+
"use_cache": false,
|
| 30 |
+
"vocab_size": 32000
|
| 31 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 1,
|
| 3 |
+
"eos_token_id": 2,
|
| 4 |
+
"max_length": 2048,
|
| 5 |
+
"pad_token_id": 0,
|
| 6 |
+
"transformers_version": "4.51.3"
|
| 7 |
+
}
|
loss.txt
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
699,0.00011922841804334894
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:06108340f68c8293f1e1c72ee69471334451e8f4aaf054ca0876b73113462a3f
|
| 3 |
+
size 2200119864
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "<unk>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "<unk>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
+
size 499723
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": null,
|
| 5 |
+
"added_tokens_decoder": {
|
| 6 |
+
"0": {
|
| 7 |
+
"content": "<unk>",
|
| 8 |
+
"lstrip": false,
|
| 9 |
+
"normalized": false,
|
| 10 |
+
"rstrip": false,
|
| 11 |
+
"single_word": false,
|
| 12 |
+
"special": true
|
| 13 |
+
},
|
| 14 |
+
"1": {
|
| 15 |
+
"content": "<s>",
|
| 16 |
+
"lstrip": false,
|
| 17 |
+
"normalized": false,
|
| 18 |
+
"rstrip": false,
|
| 19 |
+
"single_word": false,
|
| 20 |
+
"special": true
|
| 21 |
+
},
|
| 22 |
+
"2": {
|
| 23 |
+
"content": "</s>",
|
| 24 |
+
"lstrip": false,
|
| 25 |
+
"normalized": false,
|
| 26 |
+
"rstrip": false,
|
| 27 |
+
"single_word": false,
|
| 28 |
+
"special": true
|
| 29 |
+
}
|
| 30 |
+
},
|
| 31 |
+
"bos_token": "<s>",
|
| 32 |
+
"chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '<|user|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'system' %}\n{{ '<|system|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'assistant' %}\n{{ '<|assistant|>\n' + message['content'] + eos_token }}\n{% endif %}\n{% if loop.last and add_generation_prompt %}\n{{ '<|assistant|>' }}\n{% endif %}\n{% endfor %}",
|
| 33 |
+
"clean_up_tokenization_spaces": false,
|
| 34 |
+
"eos_token": "</s>",
|
| 35 |
+
"extra_special_tokens": {},
|
| 36 |
+
"legacy": false,
|
| 37 |
+
"model_max_length": 2048,
|
| 38 |
+
"pad_token": "<unk>",
|
| 39 |
+
"padding_side": "left",
|
| 40 |
+
"sp_model_kwargs": {},
|
| 41 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 42 |
+
"unk_token": "<unk>",
|
| 43 |
+
"use_default_system_prompt": false
|
| 44 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2183 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.9871794871794872,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 699,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.021367521367521368,
|
| 14 |
+
"grad_norm": 16.125,
|
| 15 |
+
"learning_rate": 1.7556090538745385e-06,
|
| 16 |
+
"logits/chosen": -3.5723183155059814,
|
| 17 |
+
"logits/rejected": -3.5035064220428467,
|
| 18 |
+
"logps/chosen": -41.11209487915039,
|
| 19 |
+
"logps/rejected": -79.84024047851562,
|
| 20 |
+
"loss": 0.693,
|
| 21 |
+
"rewards/accuracies": 0.3125,
|
| 22 |
+
"rewards/chosen": -0.00025137339252978563,
|
| 23 |
+
"rewards/margins": 0.0003745006979443133,
|
| 24 |
+
"rewards/rejected": -0.0006258742650970817,
|
| 25 |
+
"step": 5
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.042735042735042736,
|
| 29 |
+
"grad_norm": 16.625,
|
| 30 |
+
"learning_rate": 3.950120371217711e-06,
|
| 31 |
+
"logits/chosen": -3.5866427421569824,
|
| 32 |
+
"logits/rejected": -3.506498336791992,
|
| 33 |
+
"logps/chosen": -40.02031707763672,
|
| 34 |
+
"logps/rejected": -78.2303237915039,
|
| 35 |
+
"loss": 0.6804,
|
| 36 |
+
"rewards/accuracies": 0.8083333969116211,
|
| 37 |
+
"rewards/chosen": -0.0030814227648079395,
|
| 38 |
+
"rewards/margins": 0.025840703397989273,
|
| 39 |
+
"rewards/rejected": -0.028922125697135925,
|
| 40 |
+
"step": 10
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 0.0641025641025641,
|
| 44 |
+
"grad_norm": 14.75,
|
| 45 |
+
"learning_rate": 6.144631688560886e-06,
|
| 46 |
+
"logits/chosen": -3.5990729331970215,
|
| 47 |
+
"logits/rejected": -3.507888078689575,
|
| 48 |
+
"logps/chosen": -39.841556549072266,
|
| 49 |
+
"logps/rejected": -81.43814086914062,
|
| 50 |
+
"loss": 0.6304,
|
| 51 |
+
"rewards/accuracies": 0.987500011920929,
|
| 52 |
+
"rewards/chosen": -0.007047838531434536,
|
| 53 |
+
"rewards/margins": 0.131792351603508,
|
| 54 |
+
"rewards/rejected": -0.1388401836156845,
|
| 55 |
+
"step": 15
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 0.08547008547008547,
|
| 59 |
+
"grad_norm": 13.6875,
|
| 60 |
+
"learning_rate": 8.339143005904057e-06,
|
| 61 |
+
"logits/chosen": -3.5527968406677246,
|
| 62 |
+
"logits/rejected": -3.4919159412384033,
|
| 63 |
+
"logps/chosen": -39.75408935546875,
|
| 64 |
+
"logps/rejected": -80.72636413574219,
|
| 65 |
+
"loss": 0.5433,
|
| 66 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 67 |
+
"rewards/chosen": -0.014718407765030861,
|
| 68 |
+
"rewards/margins": 0.33949846029281616,
|
| 69 |
+
"rewards/rejected": -0.35421687364578247,
|
| 70 |
+
"step": 20
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.10683760683760683,
|
| 74 |
+
"grad_norm": 11.1875,
|
| 75 |
+
"learning_rate": 1.0533654323247232e-05,
|
| 76 |
+
"logits/chosen": -3.5133519172668457,
|
| 77 |
+
"logits/rejected": -3.472460985183716,
|
| 78 |
+
"logps/chosen": -40.00214385986328,
|
| 79 |
+
"logps/rejected": -85.55389404296875,
|
| 80 |
+
"loss": 0.396,
|
| 81 |
+
"rewards/accuracies": 0.98333340883255,
|
| 82 |
+
"rewards/chosen": -0.041852615773677826,
|
| 83 |
+
"rewards/margins": 0.7774692177772522,
|
| 84 |
+
"rewards/rejected": -0.8193218111991882,
|
| 85 |
+
"step": 25
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 0.1282051282051282,
|
| 89 |
+
"grad_norm": 7.59375,
|
| 90 |
+
"learning_rate": 1.2728165640590407e-05,
|
| 91 |
+
"logits/chosen": -3.4411749839782715,
|
| 92 |
+
"logits/rejected": -3.445669651031494,
|
| 93 |
+
"logps/chosen": -42.452301025390625,
|
| 94 |
+
"logps/rejected": -93.04366302490234,
|
| 95 |
+
"loss": 0.259,
|
| 96 |
+
"rewards/accuracies": 0.9958332777023315,
|
| 97 |
+
"rewards/chosen": -0.13290514051914215,
|
| 98 |
+
"rewards/margins": 1.4268956184387207,
|
| 99 |
+
"rewards/rejected": -1.5598008632659912,
|
| 100 |
+
"step": 30
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.14957264957264957,
|
| 104 |
+
"grad_norm": 4.25,
|
| 105 |
+
"learning_rate": 1.4922676957933578e-05,
|
| 106 |
+
"logits/chosen": -3.2215240001678467,
|
| 107 |
+
"logits/rejected": -3.2669529914855957,
|
| 108 |
+
"logps/chosen": -44.15301513671875,
|
| 109 |
+
"logps/rejected": -108.67581939697266,
|
| 110 |
+
"loss": 0.1196,
|
| 111 |
+
"rewards/accuracies": 1.0,
|
| 112 |
+
"rewards/chosen": -0.4359213709831238,
|
| 113 |
+
"rewards/margins": 2.7526631355285645,
|
| 114 |
+
"rewards/rejected": -3.1885838508605957,
|
| 115 |
+
"step": 35
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.17094017094017094,
|
| 119 |
+
"grad_norm": 2.046875,
|
| 120 |
+
"learning_rate": 1.5360556888469565e-05,
|
| 121 |
+
"logits/chosen": -2.778581142425537,
|
| 122 |
+
"logits/rejected": -2.812112331390381,
|
| 123 |
+
"logps/chosen": -52.46855926513672,
|
| 124 |
+
"logps/rejected": -143.13113403320312,
|
| 125 |
+
"loss": 0.0425,
|
| 126 |
+
"rewards/accuracies": 0.9916666746139526,
|
| 127 |
+
"rewards/chosen": -1.2584294080734253,
|
| 128 |
+
"rewards/margins": 5.229200839996338,
|
| 129 |
+
"rewards/rejected": -6.4876298904418945,
|
| 130 |
+
"step": 40
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.19230769230769232,
|
| 134 |
+
"grad_norm": 1.6875,
|
| 135 |
+
"learning_rate": 1.535640428282884e-05,
|
| 136 |
+
"logits/chosen": -2.3370110988616943,
|
| 137 |
+
"logits/rejected": -2.2969162464141846,
|
| 138 |
+
"logps/chosen": -51.32811737060547,
|
| 139 |
+
"logps/rejected": -161.91485595703125,
|
| 140 |
+
"loss": 0.0166,
|
| 141 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 142 |
+
"rewards/chosen": -1.2713468074798584,
|
| 143 |
+
"rewards/margins": 7.2617645263671875,
|
| 144 |
+
"rewards/rejected": -8.533111572265625,
|
| 145 |
+
"step": 45
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.21367521367521367,
|
| 149 |
+
"grad_norm": 9.8125,
|
| 150 |
+
"learning_rate": 1.5349059809872097e-05,
|
| 151 |
+
"logits/chosen": -1.9922136068344116,
|
| 152 |
+
"logits/rejected": -1.9316717386245728,
|
| 153 |
+
"logps/chosen": -57.087364196777344,
|
| 154 |
+
"logps/rejected": -180.17538452148438,
|
| 155 |
+
"loss": 0.0198,
|
| 156 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 157 |
+
"rewards/chosen": -1.6679397821426392,
|
| 158 |
+
"rewards/margins": 8.57298755645752,
|
| 159 |
+
"rewards/rejected": -10.240926742553711,
|
| 160 |
+
"step": 50
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.23504273504273504,
|
| 164 |
+
"grad_norm": 0.71484375,
|
| 165 |
+
"learning_rate": 1.5338527542732884e-05,
|
| 166 |
+
"logits/chosen": -1.988929033279419,
|
| 167 |
+
"logits/rejected": -1.921478271484375,
|
| 168 |
+
"logps/chosen": -58.5947380065918,
|
| 169 |
+
"logps/rejected": -196.42152404785156,
|
| 170 |
+
"loss": 0.0097,
|
| 171 |
+
"rewards/accuracies": 0.9958332777023315,
|
| 172 |
+
"rewards/chosen": -1.8699321746826172,
|
| 173 |
+
"rewards/margins": 9.876224517822266,
|
| 174 |
+
"rewards/rejected": -11.746156692504883,
|
| 175 |
+
"step": 55
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.2564102564102564,
|
| 179 |
+
"grad_norm": 3.0625,
|
| 180 |
+
"learning_rate": 1.532481332244717e-05,
|
| 181 |
+
"logits/chosen": -1.9385712146759033,
|
| 182 |
+
"logits/rejected": -1.8249061107635498,
|
| 183 |
+
"logps/chosen": -57.23170852661133,
|
| 184 |
+
"logps/rejected": -198.87722778320312,
|
| 185 |
+
"loss": 0.0046,
|
| 186 |
+
"rewards/accuracies": 1.0,
|
| 187 |
+
"rewards/chosen": -1.809206247329712,
|
| 188 |
+
"rewards/margins": 10.275887489318848,
|
| 189 |
+
"rewards/rejected": -12.08509349822998,
|
| 190 |
+
"step": 60
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.2777777777777778,
|
| 194 |
+
"grad_norm": 8.625,
|
| 195 |
+
"learning_rate": 1.5307924754713968e-05,
|
| 196 |
+
"logits/chosen": -1.939802885055542,
|
| 197 |
+
"logits/rejected": -1.8658767938613892,
|
| 198 |
+
"logps/chosen": -59.283775329589844,
|
| 199 |
+
"logps/rejected": -196.01666259765625,
|
| 200 |
+
"loss": 0.0104,
|
| 201 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 202 |
+
"rewards/chosen": -1.8633003234863281,
|
| 203 |
+
"rewards/margins": 9.977038383483887,
|
| 204 |
+
"rewards/rejected": -11.840339660644531,
|
| 205 |
+
"step": 65
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.29914529914529914,
|
| 209 |
+
"grad_norm": 0.32421875,
|
| 210 |
+
"learning_rate": 1.528787120567736e-05,
|
| 211 |
+
"logits/chosen": -2.0353214740753174,
|
| 212 |
+
"logits/rejected": -1.9403060674667358,
|
| 213 |
+
"logps/chosen": -53.66817092895508,
|
| 214 |
+
"logps/rejected": -192.67019653320312,
|
| 215 |
+
"loss": 0.0032,
|
| 216 |
+
"rewards/accuracies": 1.0,
|
| 217 |
+
"rewards/chosen": -1.436279058456421,
|
| 218 |
+
"rewards/margins": 10.137629508972168,
|
| 219 |
+
"rewards/rejected": -11.573908805847168,
|
| 220 |
+
"step": 70
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.32051282051282054,
|
| 224 |
+
"grad_norm": 0.33203125,
|
| 225 |
+
"learning_rate": 1.526466379673215e-05,
|
| 226 |
+
"logits/chosen": -2.051652669906616,
|
| 227 |
+
"logits/rejected": -1.9418226480484009,
|
| 228 |
+
"logps/chosen": -55.503456115722656,
|
| 229 |
+
"logps/rejected": -201.16795349121094,
|
| 230 |
+
"loss": 0.006,
|
| 231 |
+
"rewards/accuracies": 1.0,
|
| 232 |
+
"rewards/chosen": -1.6412267684936523,
|
| 233 |
+
"rewards/margins": 10.790639877319336,
|
| 234 |
+
"rewards/rejected": -12.431867599487305,
|
| 235 |
+
"step": 75
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.3418803418803419,
|
| 239 |
+
"grad_norm": 6.375,
|
| 240 |
+
"learning_rate": 1.5238315398356126e-05,
|
| 241 |
+
"logits/chosen": -2.026411533355713,
|
| 242 |
+
"logits/rejected": -1.8895785808563232,
|
| 243 |
+
"logps/chosen": -58.959983825683594,
|
| 244 |
+
"logps/rejected": -205.6489715576172,
|
| 245 |
+
"loss": 0.0054,
|
| 246 |
+
"rewards/accuracies": 1.0,
|
| 247 |
+
"rewards/chosen": -1.9467172622680664,
|
| 248 |
+
"rewards/margins": 11.003195762634277,
|
| 249 |
+
"rewards/rejected": -12.949913024902344,
|
| 250 |
+
"step": 80
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.36324786324786323,
|
| 254 |
+
"grad_norm": 0.032958984375,
|
| 255 |
+
"learning_rate": 1.5208840622972272e-05,
|
| 256 |
+
"logits/chosen": -2.0241971015930176,
|
| 257 |
+
"logits/rejected": -1.8603122234344482,
|
| 258 |
+
"logps/chosen": -59.0450553894043,
|
| 259 |
+
"logps/rejected": -208.4755401611328,
|
| 260 |
+
"loss": 0.023,
|
| 261 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 262 |
+
"rewards/chosen": -1.7697633504867554,
|
| 263 |
+
"rewards/margins": 11.253450393676758,
|
| 264 |
+
"rewards/rejected": -13.023214340209961,
|
| 265 |
+
"step": 85
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.38461538461538464,
|
| 269 |
+
"grad_norm": 1.203125,
|
| 270 |
+
"learning_rate": 1.5176255816844948e-05,
|
| 271 |
+
"logits/chosen": -1.9871257543563843,
|
| 272 |
+
"logits/rejected": -1.795332670211792,
|
| 273 |
+
"logps/chosen": -50.853111267089844,
|
| 274 |
+
"logps/rejected": -205.7463836669922,
|
| 275 |
+
"loss": 0.0063,
|
| 276 |
+
"rewards/accuracies": 1.0,
|
| 277 |
+
"rewards/chosen": -1.1833254098892212,
|
| 278 |
+
"rewards/margins": 11.639432907104492,
|
| 279 |
+
"rewards/rejected": -12.82275676727295,
|
| 280 |
+
"step": 90
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 0.405982905982906,
|
| 284 |
+
"grad_norm": 0.59765625,
|
| 285 |
+
"learning_rate": 1.5140579051014502e-05,
|
| 286 |
+
"logits/chosen": -1.9575443267822266,
|
| 287 |
+
"logits/rejected": -1.730380654335022,
|
| 288 |
+
"logps/chosen": -55.5305290222168,
|
| 289 |
+
"logps/rejected": -221.403564453125,
|
| 290 |
+
"loss": 0.0028,
|
| 291 |
+
"rewards/accuracies": 1.0,
|
| 292 |
+
"rewards/chosen": -1.54853355884552,
|
| 293 |
+
"rewards/margins": 12.512056350708008,
|
| 294 |
+
"rewards/rejected": -14.060590744018555,
|
| 295 |
+
"step": 95
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 0.42735042735042733,
|
| 299 |
+
"grad_norm": 0.107421875,
|
| 300 |
+
"learning_rate": 1.5101830111275334e-05,
|
| 301 |
+
"logits/chosen": -1.9054193496704102,
|
| 302 |
+
"logits/rejected": -1.7034189701080322,
|
| 303 |
+
"logps/chosen": -59.12446975708008,
|
| 304 |
+
"logps/rejected": -213.0598907470703,
|
| 305 |
+
"loss": 0.0117,
|
| 306 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 307 |
+
"rewards/chosen": -1.9219242334365845,
|
| 308 |
+
"rewards/margins": 11.585506439208984,
|
| 309 |
+
"rewards/rejected": -13.507431030273438,
|
| 310 |
+
"step": 100
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.44871794871794873,
|
| 314 |
+
"grad_norm": 3.484375,
|
| 315 |
+
"learning_rate": 1.5060030487203004e-05,
|
| 316 |
+
"logits/chosen": -1.84870183467865,
|
| 317 |
+
"logits/rejected": -1.5906012058258057,
|
| 318 |
+
"logps/chosen": -61.794334411621094,
|
| 319 |
+
"logps/rejected": -229.24472045898438,
|
| 320 |
+
"loss": 0.0058,
|
| 321 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 322 |
+
"rewards/chosen": -2.169433116912842,
|
| 323 |
+
"rewards/margins": 12.846372604370117,
|
| 324 |
+
"rewards/rejected": -15.015806198120117,
|
| 325 |
+
"step": 105
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.4700854700854701,
|
| 329 |
+
"grad_norm": 0.0888671875,
|
| 330 |
+
"learning_rate": 1.501520336023643e-05,
|
| 331 |
+
"logits/chosen": -1.8036330938339233,
|
| 332 |
+
"logits/rejected": -1.4892421960830688,
|
| 333 |
+
"logps/chosen": -63.50853729248047,
|
| 334 |
+
"logps/rejected": -225.1219940185547,
|
| 335 |
+
"loss": 0.0011,
|
| 336 |
+
"rewards/accuracies": 1.0,
|
| 337 |
+
"rewards/chosen": -2.3367650508880615,
|
| 338 |
+
"rewards/margins": 12.470510482788086,
|
| 339 |
+
"rewards/rejected": -14.807273864746094,
|
| 340 |
+
"step": 110
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.49145299145299143,
|
| 344 |
+
"grad_norm": 2.203125,
|
| 345 |
+
"learning_rate": 1.4967373590821828e-05,
|
| 346 |
+
"logits/chosen": -1.77732253074646,
|
| 347 |
+
"logits/rejected": -1.4749488830566406,
|
| 348 |
+
"logps/chosen": -69.56774139404297,
|
| 349 |
+
"logps/rejected": -229.52078247070312,
|
| 350 |
+
"loss": 0.0192,
|
| 351 |
+
"rewards/accuracies": 0.9916666746139526,
|
| 352 |
+
"rewards/chosen": -2.8969478607177734,
|
| 353 |
+
"rewards/margins": 12.228403091430664,
|
| 354 |
+
"rewards/rejected": -15.125350952148438,
|
| 355 |
+
"step": 115
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 0.5128205128205128,
|
| 359 |
+
"grad_norm": 0.08935546875,
|
| 360 |
+
"learning_rate": 1.491656770462546e-05,
|
| 361 |
+
"logits/chosen": -1.6314146518707275,
|
| 362 |
+
"logits/rejected": -1.2666215896606445,
|
| 363 |
+
"logps/chosen": -61.828941345214844,
|
| 364 |
+
"logps/rejected": -231.4860076904297,
|
| 365 |
+
"loss": 0.0026,
|
| 366 |
+
"rewards/accuracies": 1.0,
|
| 367 |
+
"rewards/chosen": -2.1970794200897217,
|
| 368 |
+
"rewards/margins": 13.084007263183594,
|
| 369 |
+
"rewards/rejected": -15.281087875366211,
|
| 370 |
+
"step": 120
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.5341880341880342,
|
| 374 |
+
"grad_norm": 9.5625,
|
| 375 |
+
"learning_rate": 1.4862813877822923e-05,
|
| 376 |
+
"logits/chosen": -1.695638656616211,
|
| 377 |
+
"logits/rejected": -1.3275301456451416,
|
| 378 |
+
"logps/chosen": -59.98823165893555,
|
| 379 |
+
"logps/rejected": -228.39599609375,
|
| 380 |
+
"loss": 0.0141,
|
| 381 |
+
"rewards/accuracies": 0.9916666746139526,
|
| 382 |
+
"rewards/chosen": -1.9660316705703735,
|
| 383 |
+
"rewards/margins": 13.168081283569336,
|
| 384 |
+
"rewards/rejected": -15.134114265441895,
|
| 385 |
+
"step": 125
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 0.5555555555555556,
|
| 389 |
+
"grad_norm": 0.2021484375,
|
| 390 |
+
"learning_rate": 1.4806141921473063e-05,
|
| 391 |
+
"logits/chosen": -1.7461655139923096,
|
| 392 |
+
"logits/rejected": -1.3385170698165894,
|
| 393 |
+
"logps/chosen": -57.775604248046875,
|
| 394 |
+
"logps/rejected": -232.9607391357422,
|
| 395 |
+
"loss": 0.006,
|
| 396 |
+
"rewards/accuracies": 1.0,
|
| 397 |
+
"rewards/chosen": -1.762024164199829,
|
| 398 |
+
"rewards/margins": 13.665751457214355,
|
| 399 |
+
"rewards/rejected": -15.427775382995605,
|
| 400 |
+
"step": 130
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 0.5769230769230769,
|
| 404 |
+
"grad_norm": 0.10791015625,
|
| 405 |
+
"learning_rate": 1.4746583264985202e-05,
|
| 406 |
+
"logits/chosen": -1.7675241231918335,
|
| 407 |
+
"logits/rejected": -1.349183201789856,
|
| 408 |
+
"logps/chosen": -59.05058670043945,
|
| 409 |
+
"logps/rejected": -236.23599243164062,
|
| 410 |
+
"loss": 0.0032,
|
| 411 |
+
"rewards/accuracies": 1.0,
|
| 412 |
+
"rewards/chosen": -1.866996169090271,
|
| 413 |
+
"rewards/margins": 13.846403121948242,
|
| 414 |
+
"rewards/rejected": -15.713399887084961,
|
| 415 |
+
"step": 135
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.5982905982905983,
|
| 419 |
+
"grad_norm": 0.029052734375,
|
| 420 |
+
"learning_rate": 1.468417093868888e-05,
|
| 421 |
+
"logits/chosen": -1.7844130992889404,
|
| 422 |
+
"logits/rejected": -1.393346905708313,
|
| 423 |
+
"logps/chosen": -55.1547966003418,
|
| 424 |
+
"logps/rejected": -230.82485961914062,
|
| 425 |
+
"loss": 0.0039,
|
| 426 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 427 |
+
"rewards/chosen": -1.6007912158966064,
|
| 428 |
+
"rewards/margins": 13.634336471557617,
|
| 429 |
+
"rewards/rejected": -15.235128402709961,
|
| 430 |
+
"step": 140
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.6196581196581197,
|
| 434 |
+
"grad_norm": 0.10888671875,
|
| 435 |
+
"learning_rate": 1.4618939555515721e-05,
|
| 436 |
+
"logits/chosen": -1.736817717552185,
|
| 437 |
+
"logits/rejected": -1.3590004444122314,
|
| 438 |
+
"logps/chosen": -57.30821990966797,
|
| 439 |
+
"logps/rejected": -228.35568237304688,
|
| 440 |
+
"loss": 0.0029,
|
| 441 |
+
"rewards/accuracies": 1.0,
|
| 442 |
+
"rewards/chosen": -1.7177435159683228,
|
| 443 |
+
"rewards/margins": 13.364781379699707,
|
| 444 |
+
"rewards/rejected": -15.08252239227295,
|
| 445 |
+
"step": 145
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 0.6410256410256411,
|
| 449 |
+
"grad_norm": 0.0126953125,
|
| 450 |
+
"learning_rate": 1.455092529180363e-05,
|
| 451 |
+
"logits/chosen": -1.767240285873413,
|
| 452 |
+
"logits/rejected": -1.3727061748504639,
|
| 453 |
+
"logps/chosen": -54.07440185546875,
|
| 454 |
+
"logps/rejected": -230.85598754882812,
|
| 455 |
+
"loss": 0.0004,
|
| 456 |
+
"rewards/accuracies": 1.0,
|
| 457 |
+
"rewards/chosen": -1.538040041923523,
|
| 458 |
+
"rewards/margins": 13.841748237609863,
|
| 459 |
+
"rewards/rejected": -15.379789352416992,
|
| 460 |
+
"step": 150
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 0.6623931623931624,
|
| 464 |
+
"grad_norm": 0.296875,
|
| 465 |
+
"learning_rate": 1.4480165867233946e-05,
|
| 466 |
+
"logits/chosen": -1.7429969310760498,
|
| 467 |
+
"logits/rejected": -1.3700876235961914,
|
| 468 |
+
"logps/chosen": -58.81968307495117,
|
| 469 |
+
"logps/rejected": -237.78775024414062,
|
| 470 |
+
"loss": 0.005,
|
| 471 |
+
"rewards/accuracies": 1.0,
|
| 472 |
+
"rewards/chosen": -1.9107345342636108,
|
| 473 |
+
"rewards/margins": 13.975476264953613,
|
| 474 |
+
"rewards/rejected": -15.886209487915039,
|
| 475 |
+
"step": 155
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 0.6837606837606838,
|
| 479 |
+
"grad_norm": 3.71875,
|
| 480 |
+
"learning_rate": 1.440670052391267e-05,
|
| 481 |
+
"logits/chosen": -1.7467654943466187,
|
| 482 |
+
"logits/rejected": -1.333251714706421,
|
| 483 |
+
"logps/chosen": -55.909812927246094,
|
| 484 |
+
"logps/rejected": -228.398681640625,
|
| 485 |
+
"loss": 0.006,
|
| 486 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 487 |
+
"rewards/chosen": -1.737717866897583,
|
| 488 |
+
"rewards/margins": 13.500417709350586,
|
| 489 |
+
"rewards/rejected": -15.238134384155273,
|
| 490 |
+
"step": 160
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 0.7051282051282052,
|
| 494 |
+
"grad_norm": 2.0625,
|
| 495 |
+
"learning_rate": 1.4330570004607398e-05,
|
| 496 |
+
"logits/chosen": -1.8055540323257446,
|
| 497 |
+
"logits/rejected": -1.365281343460083,
|
| 498 |
+
"logps/chosen": -56.362159729003906,
|
| 499 |
+
"logps/rejected": -241.47616577148438,
|
| 500 |
+
"loss": 0.0033,
|
| 501 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 502 |
+
"rewards/chosen": -1.7291532754898071,
|
| 503 |
+
"rewards/margins": 14.622825622558594,
|
| 504 |
+
"rewards/rejected": -16.351978302001953,
|
| 505 |
+
"step": 165
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.7264957264957265,
|
| 509 |
+
"grad_norm": 0.71875,
|
| 510 |
+
"learning_rate": 1.4251816530151986e-05,
|
| 511 |
+
"logits/chosen": -1.7655203342437744,
|
| 512 |
+
"logits/rejected": -1.3694696426391602,
|
| 513 |
+
"logps/chosen": -59.91992950439453,
|
| 514 |
+
"logps/rejected": -240.9622039794922,
|
| 515 |
+
"loss": 0.0035,
|
| 516 |
+
"rewards/accuracies": 1.0,
|
| 517 |
+
"rewards/chosen": -1.9651933908462524,
|
| 518 |
+
"rewards/margins": 14.319523811340332,
|
| 519 |
+
"rewards/rejected": -16.284717559814453,
|
| 520 |
+
"step": 170
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.7478632478632479,
|
| 524 |
+
"grad_norm": 0.228515625,
|
| 525 |
+
"learning_rate": 1.4170483776031526e-05,
|
| 526 |
+
"logits/chosen": -1.705065369606018,
|
| 527 |
+
"logits/rejected": -1.2570855617523193,
|
| 528 |
+
"logps/chosen": -59.033119201660156,
|
| 529 |
+
"logps/rejected": -244.8512725830078,
|
| 530 |
+
"loss": 0.0004,
|
| 531 |
+
"rewards/accuracies": 1.0,
|
| 532 |
+
"rewards/chosen": -1.9400125741958618,
|
| 533 |
+
"rewards/margins": 14.696588516235352,
|
| 534 |
+
"rewards/rejected": -16.636600494384766,
|
| 535 |
+
"step": 175
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.7692307692307693,
|
| 539 |
+
"grad_norm": 0.048095703125,
|
| 540 |
+
"learning_rate": 1.4086616848160574e-05,
|
| 541 |
+
"logits/chosen": -1.6960264444351196,
|
| 542 |
+
"logits/rejected": -1.2389576435089111,
|
| 543 |
+
"logps/chosen": -62.79465866088867,
|
| 544 |
+
"logps/rejected": -241.3025360107422,
|
| 545 |
+
"loss": 0.0004,
|
| 546 |
+
"rewards/accuracies": 1.0,
|
| 547 |
+
"rewards/chosen": -2.3021397590637207,
|
| 548 |
+
"rewards/margins": 14.0509614944458,
|
| 549 |
+
"rewards/rejected": -16.353099822998047,
|
| 550 |
+
"step": 180
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.7905982905982906,
|
| 554 |
+
"grad_norm": 0.087890625,
|
| 555 |
+
"learning_rate": 1.4000262257868096e-05,
|
| 556 |
+
"logits/chosen": -1.665123701095581,
|
| 557 |
+
"logits/rejected": -1.1488758325576782,
|
| 558 |
+
"logps/chosen": -61.19834518432617,
|
| 559 |
+
"logps/rejected": -244.506103515625,
|
| 560 |
+
"loss": 0.0027,
|
| 561 |
+
"rewards/accuracies": 1.0,
|
| 562 |
+
"rewards/chosen": -2.189235210418701,
|
| 563 |
+
"rewards/margins": 14.604774475097656,
|
| 564 |
+
"rewards/rejected": -16.794010162353516,
|
| 565 |
+
"step": 185
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 0.811965811965812,
|
| 569 |
+
"grad_norm": 0.921875,
|
| 570 |
+
"learning_rate": 1.3911467896102994e-05,
|
| 571 |
+
"logits/chosen": -1.6246988773345947,
|
| 572 |
+
"logits/rejected": -1.1719400882720947,
|
| 573 |
+
"logps/chosen": -58.097412109375,
|
| 574 |
+
"logps/rejected": -242.1889190673828,
|
| 575 |
+
"loss": 0.0049,
|
| 576 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 577 |
+
"rewards/chosen": -1.8747377395629883,
|
| 578 |
+
"rewards/margins": 14.691726684570312,
|
| 579 |
+
"rewards/rejected": -16.566465377807617,
|
| 580 |
+
"step": 190
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 0.8333333333333334,
|
| 584 |
+
"grad_norm": 0.216796875,
|
| 585 |
+
"learning_rate": 1.3820283006874503e-05,
|
| 586 |
+
"logits/chosen": -1.6263649463653564,
|
| 587 |
+
"logits/rejected": -1.1733753681182861,
|
| 588 |
+
"logps/chosen": -64.73831939697266,
|
| 589 |
+
"logps/rejected": -246.49752807617188,
|
| 590 |
+
"loss": 0.0091,
|
| 591 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 592 |
+
"rewards/chosen": -2.265516996383667,
|
| 593 |
+
"rewards/margins": 14.604884147644043,
|
| 594 |
+
"rewards/rejected": -16.87040138244629,
|
| 595 |
+
"step": 195
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 0.8547008547008547,
|
| 599 |
+
"grad_norm": 0.1875,
|
| 600 |
+
"learning_rate": 1.372675815994221e-05,
|
| 601 |
+
"logits/chosen": -1.5483165979385376,
|
| 602 |
+
"logits/rejected": -1.0878021717071533,
|
| 603 |
+
"logps/chosen": -52.65502166748047,
|
| 604 |
+
"logps/rejected": -248.62643432617188,
|
| 605 |
+
"loss": 0.0008,
|
| 606 |
+
"rewards/accuracies": 1.0,
|
| 607 |
+
"rewards/chosen": -1.5096328258514404,
|
| 608 |
+
"rewards/margins": 15.69578742980957,
|
| 609 |
+
"rewards/rejected": -17.205419540405273,
|
| 610 |
+
"step": 200
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 0.8760683760683761,
|
| 614 |
+
"grad_norm": 9.5,
|
| 615 |
+
"learning_rate": 1.3630945222770829e-05,
|
| 616 |
+
"logits/chosen": -1.5609447956085205,
|
| 617 |
+
"logits/rejected": -1.0415008068084717,
|
| 618 |
+
"logps/chosen": -58.0806884765625,
|
| 619 |
+
"logps/rejected": -255.28414916992188,
|
| 620 |
+
"loss": 0.0182,
|
| 621 |
+
"rewards/accuracies": 0.987500011920929,
|
| 622 |
+
"rewards/chosen": -1.7902355194091797,
|
| 623 |
+
"rewards/margins": 15.908655166625977,
|
| 624 |
+
"rewards/rejected": -17.69889259338379,
|
| 625 |
+
"step": 205
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.8974358974358975,
|
| 629 |
+
"grad_norm": 1.6484375,
|
| 630 |
+
"learning_rate": 1.3532897331765301e-05,
|
| 631 |
+
"logits/chosen": -1.5497121810913086,
|
| 632 |
+
"logits/rejected": -1.0044095516204834,
|
| 633 |
+
"logps/chosen": -59.89592742919922,
|
| 634 |
+
"logps/rejected": -257.2408752441406,
|
| 635 |
+
"loss": 0.0162,
|
| 636 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 637 |
+
"rewards/chosen": -1.991355538368225,
|
| 638 |
+
"rewards/margins": 15.942468643188477,
|
| 639 |
+
"rewards/rejected": -17.933826446533203,
|
| 640 |
+
"step": 210
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.9188034188034188,
|
| 644 |
+
"grad_norm": 0.1884765625,
|
| 645 |
+
"learning_rate": 1.3432668862802134e-05,
|
| 646 |
+
"logits/chosen": -1.4721449613571167,
|
| 647 |
+
"logits/rejected": -0.9380281567573547,
|
| 648 |
+
"logps/chosen": -59.00041961669922,
|
| 649 |
+
"logps/rejected": -255.982421875,
|
| 650 |
+
"loss": 0.0003,
|
| 651 |
+
"rewards/accuracies": 1.0,
|
| 652 |
+
"rewards/chosen": -1.8558237552642822,
|
| 653 |
+
"rewards/margins": 15.840028762817383,
|
| 654 |
+
"rewards/rejected": -17.695852279663086,
|
| 655 |
+
"step": 215
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 0.9401709401709402,
|
| 659 |
+
"grad_norm": 0.1318359375,
|
| 660 |
+
"learning_rate": 1.3330315401073371e-05,
|
| 661 |
+
"logits/chosen": -1.4841619729995728,
|
| 662 |
+
"logits/rejected": -0.9569031596183777,
|
| 663 |
+
"logps/chosen": -60.60542678833008,
|
| 664 |
+
"logps/rejected": -249.15353393554688,
|
| 665 |
+
"loss": 0.0015,
|
| 666 |
+
"rewards/accuracies": 1.0,
|
| 667 |
+
"rewards/chosen": -1.9671761989593506,
|
| 668 |
+
"rewards/margins": 15.184603691101074,
|
| 669 |
+
"rewards/rejected": -17.151779174804688,
|
| 670 |
+
"step": 220
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 0.9615384615384616,
|
| 674 |
+
"grad_norm": 9.8125,
|
| 675 |
+
"learning_rate": 1.3225893710259887e-05,
|
| 676 |
+
"logits/chosen": -1.1225636005401611,
|
| 677 |
+
"logits/rejected": -0.5225231051445007,
|
| 678 |
+
"logps/chosen": -63.623390197753906,
|
| 679 |
+
"logps/rejected": -252.66281127929688,
|
| 680 |
+
"loss": 0.0095,
|
| 681 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 682 |
+
"rewards/chosen": -2.4430124759674072,
|
| 683 |
+
"rewards/margins": 15.111320495605469,
|
| 684 |
+
"rewards/rejected": -17.554332733154297,
|
| 685 |
+
"step": 225
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 0.9829059829059829,
|
| 689 |
+
"grad_norm": 12.875,
|
| 690 |
+
"learning_rate": 1.3119461701051105e-05,
|
| 691 |
+
"logits/chosen": -1.2154924869537354,
|
| 692 |
+
"logits/rejected": -0.7296081781387329,
|
| 693 |
+
"logps/chosen": -65.45369720458984,
|
| 694 |
+
"logps/rejected": -248.72622680664062,
|
| 695 |
+
"loss": 0.0116,
|
| 696 |
+
"rewards/accuracies": 0.9916666746139526,
|
| 697 |
+
"rewards/chosen": -2.7806317806243896,
|
| 698 |
+
"rewards/margins": 14.623510360717773,
|
| 699 |
+
"rewards/rejected": -17.40414047241211,
|
| 700 |
+
"step": 230
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 0.9957264957264957,
|
| 704 |
+
"eval_logits/chosen": -1.554829716682434,
|
| 705 |
+
"eval_logits/rejected": -0.9467828869819641,
|
| 706 |
+
"eval_logps/chosen": -59.39779281616211,
|
| 707 |
+
"eval_logps/rejected": -258.1591491699219,
|
| 708 |
+
"eval_loss": 0.0002499581314623356,
|
| 709 |
+
"eval_rewards/accuracies": 1.0,
|
| 710 |
+
"eval_rewards/chosen": -1.995406150817871,
|
| 711 |
+
"eval_rewards/margins": 16.018821716308594,
|
| 712 |
+
"eval_rewards/rejected": -18.01422882080078,
|
| 713 |
+
"eval_runtime": 9.7352,
|
| 714 |
+
"eval_samples_per_second": 20.544,
|
| 715 |
+
"eval_steps_per_second": 20.544,
|
| 716 |
+
"step": 233
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.0042735042735043,
|
| 720 |
+
"grad_norm": 0.030029296875,
|
| 721 |
+
"learning_rate": 1.3011078399028605e-05,
|
| 722 |
+
"logits/chosen": -1.5108505487442017,
|
| 723 |
+
"logits/rejected": -0.9396005868911743,
|
| 724 |
+
"logps/chosen": -60.448936462402344,
|
| 725 |
+
"logps/rejected": -262.9476013183594,
|
| 726 |
+
"loss": 0.0035,
|
| 727 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 728 |
+
"rewards/chosen": -2.0872457027435303,
|
| 729 |
+
"rewards/margins": 16.35396957397461,
|
| 730 |
+
"rewards/rejected": -18.44121742248535,
|
| 731 |
+
"step": 235
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.0256410256410255,
|
| 735 |
+
"grad_norm": 0.1025390625,
|
| 736 |
+
"learning_rate": 1.2900803911931431e-05,
|
| 737 |
+
"logits/chosen": -1.5595149993896484,
|
| 738 |
+
"logits/rejected": -0.9638253450393677,
|
| 739 |
+
"logps/chosen": -56.65764617919922,
|
| 740 |
+
"logps/rejected": -258.0482482910156,
|
| 741 |
+
"loss": 0.0001,
|
| 742 |
+
"rewards/accuracies": 1.0,
|
| 743 |
+
"rewards/chosen": -1.6903537511825562,
|
| 744 |
+
"rewards/margins": 16.261157989501953,
|
| 745 |
+
"rewards/rejected": -17.95151138305664,
|
| 746 |
+
"step": 240
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.047008547008547,
|
| 750 |
+
"grad_norm": 0.022705078125,
|
| 751 |
+
"learning_rate": 1.2788699396321252e-05,
|
| 752 |
+
"logits/chosen": -1.520930290222168,
|
| 753 |
+
"logits/rejected": -0.9344191551208496,
|
| 754 |
+
"logps/chosen": -54.35309982299805,
|
| 755 |
+
"logps/rejected": -249.8235321044922,
|
| 756 |
+
"loss": 0.0002,
|
| 757 |
+
"rewards/accuracies": 1.0,
|
| 758 |
+
"rewards/chosen": -1.5775222778320312,
|
| 759 |
+
"rewards/margins": 15.838638305664062,
|
| 760 |
+
"rewards/rejected": -17.416160583496094,
|
| 761 |
+
"step": 245
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 1.0683760683760684,
|
| 765 |
+
"grad_norm": 0.00031280517578125,
|
| 766 |
+
"learning_rate": 1.2674827023665853e-05,
|
| 767 |
+
"logits/chosen": -1.5657716989517212,
|
| 768 |
+
"logits/rejected": -0.9992179870605469,
|
| 769 |
+
"logps/chosen": -57.8154411315918,
|
| 770 |
+
"logps/rejected": -263.98992919921875,
|
| 771 |
+
"loss": 0.0004,
|
| 772 |
+
"rewards/accuracies": 1.0,
|
| 773 |
+
"rewards/chosen": -1.8247381448745728,
|
| 774 |
+
"rewards/margins": 16.810840606689453,
|
| 775 |
+
"rewards/rejected": -18.635578155517578,
|
| 776 |
+
"step": 250
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 1.0897435897435896,
|
| 780 |
+
"grad_norm": 0.1005859375,
|
| 781 |
+
"learning_rate": 1.255924994585978e-05,
|
| 782 |
+
"logits/chosen": -1.532013177871704,
|
| 783 |
+
"logits/rejected": -0.9263051152229309,
|
| 784 |
+
"logps/chosen": -60.24767303466797,
|
| 785 |
+
"logps/rejected": -260.04595947265625,
|
| 786 |
+
"loss": 0.0001,
|
| 787 |
+
"rewards/accuracies": 1.0,
|
| 788 |
+
"rewards/chosen": -2.0023863315582275,
|
| 789 |
+
"rewards/margins": 16.219375610351562,
|
| 790 |
+
"rewards/rejected": -18.221763610839844,
|
| 791 |
+
"step": 255
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 1.1111111111111112,
|
| 795 |
+
"grad_norm": 0.0054931640625,
|
| 796 |
+
"learning_rate": 1.2442032260201255e-05,
|
| 797 |
+
"logits/chosen": -1.5195871591567993,
|
| 798 |
+
"logits/rejected": -0.9636222124099731,
|
| 799 |
+
"logps/chosen": -59.14739990234375,
|
| 800 |
+
"logps/rejected": -256.98883056640625,
|
| 801 |
+
"loss": 0.0001,
|
| 802 |
+
"rewards/accuracies": 1.0,
|
| 803 |
+
"rewards/chosen": -1.9601608514785767,
|
| 804 |
+
"rewards/margins": 15.914080619812012,
|
| 805 |
+
"rewards/rejected": -17.874242782592773,
|
| 806 |
+
"step": 260
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 1.1324786324786325,
|
| 810 |
+
"grad_norm": 0.283203125,
|
| 811 |
+
"learning_rate": 1.2323238973844796e-05,
|
| 812 |
+
"logits/chosen": -1.5553245544433594,
|
| 813 |
+
"logits/rejected": -0.9653908014297485,
|
| 814 |
+
"logps/chosen": -59.334739685058594,
|
| 815 |
+
"logps/rejected": -266.408935546875,
|
| 816 |
+
"loss": 0.0002,
|
| 817 |
+
"rewards/accuracies": 1.0,
|
| 818 |
+
"rewards/chosen": -1.896758794784546,
|
| 819 |
+
"rewards/margins": 16.70488739013672,
|
| 820 |
+
"rewards/rejected": -18.601646423339844,
|
| 821 |
+
"step": 265
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.1538461538461537,
|
| 825 |
+
"grad_norm": 0.0145263671875,
|
| 826 |
+
"learning_rate": 1.2202935967749212e-05,
|
| 827 |
+
"logits/chosen": -1.5117428302764893,
|
| 828 |
+
"logits/rejected": -0.9256662130355835,
|
| 829 |
+
"logps/chosen": -62.41327667236328,
|
| 830 |
+
"logps/rejected": -261.29559326171875,
|
| 831 |
+
"loss": 0.0006,
|
| 832 |
+
"rewards/accuracies": 1.0,
|
| 833 |
+
"rewards/chosen": -2.1235432624816895,
|
| 834 |
+
"rewards/margins": 16.034034729003906,
|
| 835 |
+
"rewards/rejected": -18.15757942199707,
|
| 836 |
+
"step": 270
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.1752136752136753,
|
| 840 |
+
"grad_norm": 0.16796875,
|
| 841 |
+
"learning_rate": 1.2081189960141038e-05,
|
| 842 |
+
"logits/chosen": -1.5283492803573608,
|
| 843 |
+
"logits/rejected": -0.9296523332595825,
|
| 844 |
+
"logps/chosen": -57.735595703125,
|
| 845 |
+
"logps/rejected": -260.36627197265625,
|
| 846 |
+
"loss": 0.0049,
|
| 847 |
+
"rewards/accuracies": 0.9958332777023315,
|
| 848 |
+
"rewards/chosen": -1.738776445388794,
|
| 849 |
+
"rewards/margins": 16.458282470703125,
|
| 850 |
+
"rewards/rejected": -18.19706153869629,
|
| 851 |
+
"step": 275
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 1.1965811965811965,
|
| 855 |
+
"grad_norm": 0.51171875,
|
| 856 |
+
"learning_rate": 1.1958068469513604e-05,
|
| 857 |
+
"logits/chosen": -1.5641071796417236,
|
| 858 |
+
"logits/rejected": -0.977310299873352,
|
| 859 |
+
"logps/chosen": -59.65924072265625,
|
| 860 |
+
"logps/rejected": -269.66253662109375,
|
| 861 |
+
"loss": 0.0029,
|
| 862 |
+
"rewards/accuracies": 1.0,
|
| 863 |
+
"rewards/chosen": -1.9156348705291748,
|
| 864 |
+
"rewards/margins": 17.137256622314453,
|
| 865 |
+
"rewards/rejected": -19.05289077758789,
|
| 866 |
+
"step": 280
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.217948717948718,
|
| 870 |
+
"grad_norm": 0.09033203125,
|
| 871 |
+
"learning_rate": 1.1833639777182316e-05,
|
| 872 |
+
"logits/chosen": -1.5035396814346313,
|
| 873 |
+
"logits/rejected": -0.8975561857223511,
|
| 874 |
+
"logps/chosen": -54.24858856201172,
|
| 875 |
+
"logps/rejected": -262.6181945800781,
|
| 876 |
+
"loss": 0.0001,
|
| 877 |
+
"rewards/accuracies": 1.0,
|
| 878 |
+
"rewards/chosen": -1.5752283334732056,
|
| 879 |
+
"rewards/margins": 16.97848129272461,
|
| 880 |
+
"rewards/rejected": -18.5537109375,
|
| 881 |
+
"step": 285
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 1.2393162393162394,
|
| 885 |
+
"grad_norm": 0.025390625,
|
| 886 |
+
"learning_rate": 1.170797288941685e-05,
|
| 887 |
+
"logits/chosen": -1.4596306085586548,
|
| 888 |
+
"logits/rejected": -0.8935238718986511,
|
| 889 |
+
"logps/chosen": -60.20256805419922,
|
| 890 |
+
"logps/rejected": -270.41259765625,
|
| 891 |
+
"loss": 0.001,
|
| 892 |
+
"rewards/accuracies": 1.0,
|
| 893 |
+
"rewards/chosen": -2.0154104232788086,
|
| 894 |
+
"rewards/margins": 16.990856170654297,
|
| 895 |
+
"rewards/rejected": -19.00626564025879,
|
| 896 |
+
"step": 290
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 1.2606837606837606,
|
| 900 |
+
"grad_norm": 0.443359375,
|
| 901 |
+
"learning_rate": 1.1581137499171342e-05,
|
| 902 |
+
"logits/chosen": -1.478342890739441,
|
| 903 |
+
"logits/rejected": -0.8745730519294739,
|
| 904 |
+
"logps/chosen": -61.747718811035156,
|
| 905 |
+
"logps/rejected": -260.48358154296875,
|
| 906 |
+
"loss": 0.0006,
|
| 907 |
+
"rewards/accuracies": 1.0,
|
| 908 |
+
"rewards/chosen": -2.1923012733459473,
|
| 909 |
+
"rewards/margins": 16.068017959594727,
|
| 910 |
+
"rewards/rejected": -18.260318756103516,
|
| 911 |
+
"step": 295
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 1.282051282051282,
|
| 915 |
+
"grad_norm": 0.0164794921875,
|
| 916 |
+
"learning_rate": 1.145320394743371e-05,
|
| 917 |
+
"logits/chosen": -1.439082384109497,
|
| 918 |
+
"logits/rejected": -0.8525334596633911,
|
| 919 |
+
"logps/chosen": -59.48680877685547,
|
| 920 |
+
"logps/rejected": -256.1888122558594,
|
| 921 |
+
"loss": 0.0002,
|
| 922 |
+
"rewards/accuracies": 1.0,
|
| 923 |
+
"rewards/chosen": -1.9997012615203857,
|
| 924 |
+
"rewards/margins": 16.03377914428711,
|
| 925 |
+
"rewards/rejected": -18.033479690551758,
|
| 926 |
+
"step": 300
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.3034188034188035,
|
| 930 |
+
"grad_norm": 0.004180908203125,
|
| 931 |
+
"learning_rate": 1.1324243184215622e-05,
|
| 932 |
+
"logits/chosen": -1.415179967880249,
|
| 933 |
+
"logits/rejected": -0.811850368976593,
|
| 934 |
+
"logps/chosen": -60.232139587402344,
|
| 935 |
+
"logps/rejected": -271.9323425292969,
|
| 936 |
+
"loss": 0.0,
|
| 937 |
+
"rewards/accuracies": 1.0,
|
| 938 |
+
"rewards/chosen": -2.1755707263946533,
|
| 939 |
+
"rewards/margins": 17.213939666748047,
|
| 940 |
+
"rewards/rejected": -19.389511108398438,
|
| 941 |
+
"step": 305
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.3247863247863247,
|
| 945 |
+
"grad_norm": 0.0079345703125,
|
| 946 |
+
"learning_rate": 1.1194326729204686e-05,
|
| 947 |
+
"logits/chosen": -1.4362504482269287,
|
| 948 |
+
"logits/rejected": -0.8141258358955383,
|
| 949 |
+
"logps/chosen": -60.56780242919922,
|
| 950 |
+
"logps/rejected": -269.37628173828125,
|
| 951 |
+
"loss": 0.0002,
|
| 952 |
+
"rewards/accuracies": 1.0,
|
| 953 |
+
"rewards/chosen": -2.1700046062469482,
|
| 954 |
+
"rewards/margins": 17.103158950805664,
|
| 955 |
+
"rewards/rejected": -19.273160934448242,
|
| 956 |
+
"step": 310
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.3461538461538463,
|
| 960 |
+
"grad_norm": 0.0264892578125,
|
| 961 |
+
"learning_rate": 1.1063526632100717e-05,
|
| 962 |
+
"logits/chosen": -1.4126843214035034,
|
| 963 |
+
"logits/rejected": -0.8487581014633179,
|
| 964 |
+
"logps/chosen": -63.435699462890625,
|
| 965 |
+
"logps/rejected": -263.54461669921875,
|
| 966 |
+
"loss": 0.0035,
|
| 967 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 968 |
+
"rewards/chosen": -2.3221802711486816,
|
| 969 |
+
"rewards/margins": 16.341337203979492,
|
| 970 |
+
"rewards/rejected": -18.66351890563965,
|
| 971 |
+
"step": 315
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 1.3675213675213675,
|
| 975 |
+
"grad_norm": 0.01202392578125,
|
| 976 |
+
"learning_rate": 1.0931915432658055e-05,
|
| 977 |
+
"logits/chosen": -1.3911561965942383,
|
| 978 |
+
"logits/rejected": -0.7923445701599121,
|
| 979 |
+
"logps/chosen": -63.30759811401367,
|
| 980 |
+
"logps/rejected": -268.752685546875,
|
| 981 |
+
"loss": 0.0001,
|
| 982 |
+
"rewards/accuracies": 1.0,
|
| 983 |
+
"rewards/chosen": -2.365898609161377,
|
| 984 |
+
"rewards/margins": 16.713449478149414,
|
| 985 |
+
"rewards/rejected": -19.079347610473633,
|
| 986 |
+
"step": 320
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 1.3888888888888888,
|
| 990 |
+
"grad_norm": 0.01446533203125,
|
| 991 |
+
"learning_rate": 1.0799566120456133e-05,
|
| 992 |
+
"logits/chosen": -1.3831051588058472,
|
| 993 |
+
"logits/rejected": -0.7522771954536438,
|
| 994 |
+
"logps/chosen": -64.623046875,
|
| 995 |
+
"logps/rejected": -268.64984130859375,
|
| 996 |
+
"loss": 0.0002,
|
| 997 |
+
"rewards/accuracies": 1.0,
|
| 998 |
+
"rewards/chosen": -2.4689812660217285,
|
| 999 |
+
"rewards/margins": 16.626056671142578,
|
| 1000 |
+
"rewards/rejected": -19.09503746032715,
|
| 1001 |
+
"step": 325
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 1.4102564102564101,
|
| 1005 |
+
"grad_norm": 0.01055908203125,
|
| 1006 |
+
"learning_rate": 1.066655209442054e-05,
|
| 1007 |
+
"logits/chosen": -1.4041249752044678,
|
| 1008 |
+
"logits/rejected": -0.7933310866355896,
|
| 1009 |
+
"logps/chosen": -65.8058853149414,
|
| 1010 |
+
"logps/rejected": -272.6845397949219,
|
| 1011 |
+
"loss": 0.0021,
|
| 1012 |
+
"rewards/accuracies": 1.0,
|
| 1013 |
+
"rewards/chosen": -2.579711675643921,
|
| 1014 |
+
"rewards/margins": 16.771991729736328,
|
| 1015 |
+
"rewards/rejected": -19.351699829101562,
|
| 1016 |
+
"step": 330
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 1.4316239316239316,
|
| 1020 |
+
"grad_norm": 0.0155029296875,
|
| 1021 |
+
"learning_rate": 1.0532947122117101e-05,
|
| 1022 |
+
"logits/chosen": -1.3893976211547852,
|
| 1023 |
+
"logits/rejected": -0.7593523263931274,
|
| 1024 |
+
"logps/chosen": -65.74365234375,
|
| 1025 |
+
"logps/rejected": -271.69537353515625,
|
| 1026 |
+
"loss": 0.0,
|
| 1027 |
+
"rewards/accuracies": 1.0,
|
| 1028 |
+
"rewards/chosen": -2.5910518169403076,
|
| 1029 |
+
"rewards/margins": 16.82119369506836,
|
| 1030 |
+
"rewards/rejected": -19.41224479675293,
|
| 1031 |
+
"step": 335
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.452991452991453,
|
| 1035 |
+
"grad_norm": 0.0167236328125,
|
| 1036 |
+
"learning_rate": 1.0398825298841499e-05,
|
| 1037 |
+
"logits/chosen": -1.3964606523513794,
|
| 1038 |
+
"logits/rejected": -0.8304880261421204,
|
| 1039 |
+
"logps/chosen": -69.43986511230469,
|
| 1040 |
+
"logps/rejected": -275.15960693359375,
|
| 1041 |
+
"loss": 0.0003,
|
| 1042 |
+
"rewards/accuracies": 1.0,
|
| 1043 |
+
"rewards/chosen": -2.830099582672119,
|
| 1044 |
+
"rewards/margins": 16.80275535583496,
|
| 1045 |
+
"rewards/rejected": -19.632854461669922,
|
| 1046 |
+
"step": 340
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.4743589743589745,
|
| 1050 |
+
"grad_norm": 0.0263671875,
|
| 1051 |
+
"learning_rate": 1.0264261006527144e-05,
|
| 1052 |
+
"logits/chosen": -1.4159036874771118,
|
| 1053 |
+
"logits/rejected": -0.8477792739868164,
|
| 1054 |
+
"logps/chosen": -66.49871063232422,
|
| 1055 |
+
"logps/rejected": -265.2044372558594,
|
| 1056 |
+
"loss": 0.004,
|
| 1057 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 1058 |
+
"rewards/chosen": -2.641693592071533,
|
| 1059 |
+
"rewards/margins": 16.258649826049805,
|
| 1060 |
+
"rewards/rejected": -18.900344848632812,
|
| 1061 |
+
"step": 345
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 1.4957264957264957,
|
| 1065 |
+
"grad_norm": 0.4140625,
|
| 1066 |
+
"learning_rate": 1.0129328872494075e-05,
|
| 1067 |
+
"logits/chosen": -1.4811489582061768,
|
| 1068 |
+
"logits/rejected": -0.9099611043930054,
|
| 1069 |
+
"logps/chosen": -66.62818908691406,
|
| 1070 |
+
"logps/rejected": -271.5672912597656,
|
| 1071 |
+
"loss": 0.003,
|
| 1072 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 1073 |
+
"rewards/chosen": -2.69557523727417,
|
| 1074 |
+
"rewards/margins": 16.735515594482422,
|
| 1075 |
+
"rewards/rejected": -19.431089401245117,
|
| 1076 |
+
"step": 350
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 1.517094017094017,
|
| 1080 |
+
"grad_norm": 0.134765625,
|
| 1081 |
+
"learning_rate": 9.994103728061786e-06,
|
| 1082 |
+
"logits/chosen": -1.434282660484314,
|
| 1083 |
+
"logits/rejected": -0.8590477705001831,
|
| 1084 |
+
"logps/chosen": -68.98523712158203,
|
| 1085 |
+
"logps/rejected": -271.1317138671875,
|
| 1086 |
+
"loss": 0.0003,
|
| 1087 |
+
"rewards/accuracies": 1.0,
|
| 1088 |
+
"rewards/chosen": -2.879164934158325,
|
| 1089 |
+
"rewards/margins": 16.430065155029297,
|
| 1090 |
+
"rewards/rejected": -19.30923080444336,
|
| 1091 |
+
"step": 355
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 1.5384615384615383,
|
| 1095 |
+
"grad_norm": 0.00104522705078125,
|
| 1096 |
+
"learning_rate": 9.858660567048902e-06,
|
| 1097 |
+
"logits/chosen": -1.4451745748519897,
|
| 1098 |
+
"logits/rejected": -0.8094388842582703,
|
| 1099 |
+
"logps/chosen": -71.38107299804688,
|
| 1100 |
+
"logps/rejected": -284.3163757324219,
|
| 1101 |
+
"loss": 0.0,
|
| 1102 |
+
"rewards/accuracies": 1.0,
|
| 1103 |
+
"rewards/chosen": -3.1282448768615723,
|
| 1104 |
+
"rewards/margins": 17.431547164916992,
|
| 1105 |
+
"rewards/rejected": -20.559795379638672,
|
| 1106 |
+
"step": 360
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 1.5598290598290598,
|
| 1110 |
+
"grad_norm": 0.06591796875,
|
| 1111 |
+
"learning_rate": 9.72307450418274e-06,
|
| 1112 |
+
"logits/chosen": -1.4680107831954956,
|
| 1113 |
+
"logits/rejected": -0.8780597448348999,
|
| 1114 |
+
"logps/chosen": -66.4084701538086,
|
| 1115 |
+
"logps/rejected": -269.15948486328125,
|
| 1116 |
+
"loss": 0.0004,
|
| 1117 |
+
"rewards/accuracies": 1.0,
|
| 1118 |
+
"rewards/chosen": -2.635270357131958,
|
| 1119 |
+
"rewards/margins": 16.57504653930664,
|
| 1120 |
+
"rewards/rejected": -19.210315704345703,
|
| 1121 |
+
"step": 365
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 1.5811965811965814,
|
| 1125 |
+
"grad_norm": 1.921875,
|
| 1126 |
+
"learning_rate": 9.587420733441835e-06,
|
| 1127 |
+
"logits/chosen": -1.484803557395935,
|
| 1128 |
+
"logits/rejected": -0.9129883646965027,
|
| 1129 |
+
"logps/chosen": -63.17878341674805,
|
| 1130 |
+
"logps/rejected": -272.344482421875,
|
| 1131 |
+
"loss": 0.0006,
|
| 1132 |
+
"rewards/accuracies": 1.0,
|
| 1133 |
+
"rewards/chosen": -2.336841106414795,
|
| 1134 |
+
"rewards/margins": 17.08325958251953,
|
| 1135 |
+
"rewards/rejected": -19.420101165771484,
|
| 1136 |
+
"step": 370
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.6025641025641026,
|
| 1140 |
+
"grad_norm": 0.000766754150390625,
|
| 1141 |
+
"learning_rate": 9.45177448635447e-06,
|
| 1142 |
+
"logits/chosen": -1.4960134029388428,
|
| 1143 |
+
"logits/rejected": -0.8223481178283691,
|
| 1144 |
+
"logps/chosen": -63.178985595703125,
|
| 1145 |
+
"logps/rejected": -274.9540100097656,
|
| 1146 |
+
"loss": 0.0,
|
| 1147 |
+
"rewards/accuracies": 1.0,
|
| 1148 |
+
"rewards/chosen": -2.362715482711792,
|
| 1149 |
+
"rewards/margins": 17.505586624145508,
|
| 1150 |
+
"rewards/rejected": -19.868301391601562,
|
| 1151 |
+
"step": 375
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.623931623931624,
|
| 1155 |
+
"grad_norm": 0.046630859375,
|
| 1156 |
+
"learning_rate": 9.316210990276434e-06,
|
| 1157 |
+
"logits/chosen": -1.425362467765808,
|
| 1158 |
+
"logits/rejected": -0.851107120513916,
|
| 1159 |
+
"logps/chosen": -62.67730712890625,
|
| 1160 |
+
"logps/rejected": -263.357666015625,
|
| 1161 |
+
"loss": 0.0001,
|
| 1162 |
+
"rewards/accuracies": 1.0,
|
| 1163 |
+
"rewards/chosen": -2.2725672721862793,
|
| 1164 |
+
"rewards/margins": 16.371471405029297,
|
| 1165 |
+
"rewards/rejected": -18.644039154052734,
|
| 1166 |
+
"step": 380
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 1.6452991452991452,
|
| 1170 |
+
"grad_norm": 0.0045166015625,
|
| 1171 |
+
"learning_rate": 9.18080542667105e-06,
|
| 1172 |
+
"logits/chosen": -1.4633301496505737,
|
| 1173 |
+
"logits/rejected": -0.8226639628410339,
|
| 1174 |
+
"logps/chosen": -63.28871536254883,
|
| 1175 |
+
"logps/rejected": -280.1250915527344,
|
| 1176 |
+
"loss": 0.0001,
|
| 1177 |
+
"rewards/accuracies": 1.0,
|
| 1178 |
+
"rewards/chosen": -2.2893106937408447,
|
| 1179 |
+
"rewards/margins": 17.715099334716797,
|
| 1180 |
+
"rewards/rejected": -20.004411697387695,
|
| 1181 |
+
"step": 385
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 1.6666666666666665,
|
| 1185 |
+
"grad_norm": 0.004486083984375,
|
| 1186 |
+
"learning_rate": 9.045632889414686e-06,
|
| 1187 |
+
"logits/chosen": -1.451174020767212,
|
| 1188 |
+
"logits/rejected": -0.8335480690002441,
|
| 1189 |
+
"logps/chosen": -61.83076095581055,
|
| 1190 |
+
"logps/rejected": -272.6547546386719,
|
| 1191 |
+
"loss": 0.0001,
|
| 1192 |
+
"rewards/accuracies": 1.0,
|
| 1193 |
+
"rewards/chosen": -2.2272868156433105,
|
| 1194 |
+
"rewards/margins": 17.232776641845703,
|
| 1195 |
+
"rewards/rejected": -19.460063934326172,
|
| 1196 |
+
"step": 390
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 1.688034188034188,
|
| 1200 |
+
"grad_norm": 0.01153564453125,
|
| 1201 |
+
"learning_rate": 8.910768343150828e-06,
|
| 1202 |
+
"logits/chosen": -1.4321696758270264,
|
| 1203 |
+
"logits/rejected": -0.8404842615127563,
|
| 1204 |
+
"logps/chosen": -64.93434143066406,
|
| 1205 |
+
"logps/rejected": -273.5433044433594,
|
| 1206 |
+
"loss": 0.0001,
|
| 1207 |
+
"rewards/accuracies": 1.0,
|
| 1208 |
+
"rewards/chosen": -2.5962464809417725,
|
| 1209 |
+
"rewards/margins": 17.047863006591797,
|
| 1210 |
+
"rewards/rejected": -19.64410972595215,
|
| 1211 |
+
"step": 395
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 1.7094017094017095,
|
| 1215 |
+
"grad_norm": 0.006378173828125,
|
| 1216 |
+
"learning_rate": 8.77628658171581e-06,
|
| 1217 |
+
"logits/chosen": -1.4662706851959229,
|
| 1218 |
+
"logits/rejected": -0.8858024477958679,
|
| 1219 |
+
"logps/chosen": -63.3156852722168,
|
| 1220 |
+
"logps/rejected": -268.3216552734375,
|
| 1221 |
+
"loss": 0.0003,
|
| 1222 |
+
"rewards/accuracies": 1.0,
|
| 1223 |
+
"rewards/chosen": -2.346360683441162,
|
| 1224 |
+
"rewards/margins": 16.638729095458984,
|
| 1225 |
+
"rewards/rejected": -18.98509407043457,
|
| 1226 |
+
"step": 400
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 1.7307692307692308,
|
| 1230 |
+
"grad_norm": 0.033203125,
|
| 1231 |
+
"learning_rate": 8.642262186659298e-06,
|
| 1232 |
+
"logits/chosen": -1.4258873462677002,
|
| 1233 |
+
"logits/rejected": -0.8420296907424927,
|
| 1234 |
+
"logps/chosen": -64.47160339355469,
|
| 1235 |
+
"logps/rejected": -271.00213623046875,
|
| 1236 |
+
"loss": 0.0002,
|
| 1237 |
+
"rewards/accuracies": 1.0,
|
| 1238 |
+
"rewards/chosen": -2.43169903755188,
|
| 1239 |
+
"rewards/margins": 16.784656524658203,
|
| 1240 |
+
"rewards/rejected": -19.21635627746582,
|
| 1241 |
+
"step": 405
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.7521367521367521,
|
| 1245 |
+
"grad_norm": 0.023681640625,
|
| 1246 |
+
"learning_rate": 8.508769485882487e-06,
|
| 1247 |
+
"logits/chosen": -1.4450222253799438,
|
| 1248 |
+
"logits/rejected": -0.8718585968017578,
|
| 1249 |
+
"logps/chosen": -69.8866195678711,
|
| 1250 |
+
"logps/rejected": -272.61322021484375,
|
| 1251 |
+
"loss": 0.0022,
|
| 1252 |
+
"rewards/accuracies": 1.0,
|
| 1253 |
+
"rewards/chosen": -2.834794759750366,
|
| 1254 |
+
"rewards/margins": 16.60897445678711,
|
| 1255 |
+
"rewards/rejected": -19.443769454956055,
|
| 1256 |
+
"step": 410
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.7735042735042734,
|
| 1260 |
+
"grad_norm": 0.703125,
|
| 1261 |
+
"learning_rate": 8.375882512416969e-06,
|
| 1262 |
+
"logits/chosen": -1.4202231168746948,
|
| 1263 |
+
"logits/rejected": -0.7998637557029724,
|
| 1264 |
+
"logps/chosen": -67.1268310546875,
|
| 1265 |
+
"logps/rejected": -275.6379089355469,
|
| 1266 |
+
"loss": 0.0006,
|
| 1267 |
+
"rewards/accuracies": 1.0,
|
| 1268 |
+
"rewards/chosen": -2.6624515056610107,
|
| 1269 |
+
"rewards/margins": 17.090303421020508,
|
| 1270 |
+
"rewards/rejected": -19.752756118774414,
|
| 1271 |
+
"step": 415
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 1.7948717948717947,
|
| 1275 |
+
"grad_norm": 0.0458984375,
|
| 1276 |
+
"learning_rate": 8.243674963367137e-06,
|
| 1277 |
+
"logits/chosen": -1.420807957649231,
|
| 1278 |
+
"logits/rejected": -0.8041313886642456,
|
| 1279 |
+
"logps/chosen": -68.99574279785156,
|
| 1280 |
+
"logps/rejected": -272.53936767578125,
|
| 1281 |
+
"loss": 0.0033,
|
| 1282 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 1283 |
+
"rewards/chosen": -2.8813891410827637,
|
| 1284 |
+
"rewards/margins": 16.42134666442871,
|
| 1285 |
+
"rewards/rejected": -19.302734375,
|
| 1286 |
+
"step": 420
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 1.8162393162393162,
|
| 1290 |
+
"grad_norm": 0.000850677490234375,
|
| 1291 |
+
"learning_rate": 8.11222015903888e-06,
|
| 1292 |
+
"logits/chosen": -1.4625945091247559,
|
| 1293 |
+
"logits/rejected": -0.8097749948501587,
|
| 1294 |
+
"logps/chosen": -67.76467895507812,
|
| 1295 |
+
"logps/rejected": -278.23931884765625,
|
| 1296 |
+
"loss": 0.0032,
|
| 1297 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 1298 |
+
"rewards/chosen": -2.830801010131836,
|
| 1299 |
+
"rewards/margins": 17.239028930664062,
|
| 1300 |
+
"rewards/rejected": -20.0698299407959,
|
| 1301 |
+
"step": 425
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 1.8376068376068377,
|
| 1305 |
+
"grad_norm": 0.03515625,
|
| 1306 |
+
"learning_rate": 7.981591002277265e-06,
|
| 1307 |
+
"logits/chosen": -1.4442012310028076,
|
| 1308 |
+
"logits/rejected": -0.8179518580436707,
|
| 1309 |
+
"logps/chosen": -69.31373596191406,
|
| 1310 |
+
"logps/rejected": -273.0030822753906,
|
| 1311 |
+
"loss": 0.0002,
|
| 1312 |
+
"rewards/accuracies": 1.0,
|
| 1313 |
+
"rewards/chosen": -2.9278433322906494,
|
| 1314 |
+
"rewards/margins": 16.736169815063477,
|
| 1315 |
+
"rewards/rejected": -19.66401481628418,
|
| 1316 |
+
"step": 430
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 1.858974358974359,
|
| 1320 |
+
"grad_norm": 0.01116943359375,
|
| 1321 |
+
"learning_rate": 7.851859938035712e-06,
|
| 1322 |
+
"logits/chosen": -1.4216701984405518,
|
| 1323 |
+
"logits/rejected": -0.7942286729812622,
|
| 1324 |
+
"logps/chosen": -71.70731353759766,
|
| 1325 |
+
"logps/rejected": -282.1054382324219,
|
| 1326 |
+
"loss": 0.0001,
|
| 1327 |
+
"rewards/accuracies": 1.0,
|
| 1328 |
+
"rewards/chosen": -3.118105888366699,
|
| 1329 |
+
"rewards/margins": 17.2371768951416,
|
| 1330 |
+
"rewards/rejected": -20.355281829833984,
|
| 1331 |
+
"step": 435
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 1.8803418803418803,
|
| 1335 |
+
"grad_norm": 0.0279541015625,
|
| 1336 |
+
"learning_rate": 7.723098913199118e-06,
|
| 1337 |
+
"logits/chosen": -1.4586832523345947,
|
| 1338 |
+
"logits/rejected": -0.8305650949478149,
|
| 1339 |
+
"logps/chosen": -67.39842224121094,
|
| 1340 |
+
"logps/rejected": -269.7090148925781,
|
| 1341 |
+
"loss": 0.0001,
|
| 1342 |
+
"rewards/accuracies": 1.0,
|
| 1343 |
+
"rewards/chosen": -2.8383772373199463,
|
| 1344 |
+
"rewards/margins": 16.61200714111328,
|
| 1345 |
+
"rewards/rejected": -19.450382232666016,
|
| 1346 |
+
"step": 440
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.9017094017094016,
|
| 1350 |
+
"grad_norm": 0.0419921875,
|
| 1351 |
+
"learning_rate": 7.595379336683204e-06,
|
| 1352 |
+
"logits/chosen": -1.433342456817627,
|
| 1353 |
+
"logits/rejected": -0.7530536651611328,
|
| 1354 |
+
"logps/chosen": -65.9086685180664,
|
| 1355 |
+
"logps/rejected": -276.2018127441406,
|
| 1356 |
+
"loss": 0.0,
|
| 1357 |
+
"rewards/accuracies": 1.0,
|
| 1358 |
+
"rewards/chosen": -2.7526674270629883,
|
| 1359 |
+
"rewards/margins": 17.266189575195312,
|
| 1360 |
+
"rewards/rejected": -20.018856048583984,
|
| 1361 |
+
"step": 445
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.9230769230769231,
|
| 1365 |
+
"grad_norm": 0.00433349609375,
|
| 1366 |
+
"learning_rate": 7.468772039832218e-06,
|
| 1367 |
+
"logits/chosen": -1.4057344198226929,
|
| 1368 |
+
"logits/rejected": -0.7420554757118225,
|
| 1369 |
+
"logps/chosen": -61.406890869140625,
|
| 1370 |
+
"logps/rejected": -270.76568603515625,
|
| 1371 |
+
"loss": 0.0,
|
| 1372 |
+
"rewards/accuracies": 1.0,
|
| 1373 |
+
"rewards/chosen": -2.308377742767334,
|
| 1374 |
+
"rewards/margins": 17.16427230834961,
|
| 1375 |
+
"rewards/rejected": -19.472646713256836,
|
| 1376 |
+
"step": 450
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 1.9444444444444444,
|
| 1380 |
+
"grad_norm": 0.00634765625,
|
| 1381 |
+
"learning_rate": 7.3433472371369404e-06,
|
| 1382 |
+
"logits/chosen": -1.4746873378753662,
|
| 1383 |
+
"logits/rejected": -0.8472849726676941,
|
| 1384 |
+
"logps/chosen": -66.808837890625,
|
| 1385 |
+
"logps/rejected": -274.1222839355469,
|
| 1386 |
+
"loss": 0.0001,
|
| 1387 |
+
"rewards/accuracies": 1.0,
|
| 1388 |
+
"rewards/chosen": -2.7300994396209717,
|
| 1389 |
+
"rewards/margins": 16.893524169921875,
|
| 1390 |
+
"rewards/rejected": -19.623624801635742,
|
| 1391 |
+
"step": 455
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 1.965811965811966,
|
| 1395 |
+
"grad_norm": 0.00092315673828125,
|
| 1396 |
+
"learning_rate": 7.219174487294784e-06,
|
| 1397 |
+
"logits/chosen": -1.4693877696990967,
|
| 1398 |
+
"logits/rejected": -0.856170654296875,
|
| 1399 |
+
"logps/chosen": -64.80795288085938,
|
| 1400 |
+
"logps/rejected": -276.7772216796875,
|
| 1401 |
+
"loss": 0.0001,
|
| 1402 |
+
"rewards/accuracies": 1.0,
|
| 1403 |
+
"rewards/chosen": -2.508439779281616,
|
| 1404 |
+
"rewards/margins": 17.357221603393555,
|
| 1405 |
+
"rewards/rejected": -19.865665435791016,
|
| 1406 |
+
"step": 460
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 1.9871794871794872,
|
| 1410 |
+
"grad_norm": 0.0234375,
|
| 1411 |
+
"learning_rate": 7.0963226546336e-06,
|
| 1412 |
+
"logits/chosen": -1.475427269935608,
|
| 1413 |
+
"logits/rejected": -0.8358310461044312,
|
| 1414 |
+
"logps/chosen": -66.6759262084961,
|
| 1415 |
+
"logps/rejected": -272.0386962890625,
|
| 1416 |
+
"loss": 0.0001,
|
| 1417 |
+
"rewards/accuracies": 1.0,
|
| 1418 |
+
"rewards/chosen": -2.7268948554992676,
|
| 1419 |
+
"rewards/margins": 16.760303497314453,
|
| 1420 |
+
"rewards/rejected": -19.487197875976562,
|
| 1421 |
+
"step": 465
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 1.9914529914529915,
|
| 1425 |
+
"eval_logits/chosen": -1.4994289875030518,
|
| 1426 |
+
"eval_logits/rejected": -0.8360934257507324,
|
| 1427 |
+
"eval_logps/chosen": -67.28651428222656,
|
| 1428 |
+
"eval_logps/rejected": -275.5657043457031,
|
| 1429 |
+
"eval_loss": 0.00012571073602885008,
|
| 1430 |
+
"eval_rewards/accuracies": 1.0,
|
| 1431 |
+
"eval_rewards/chosen": -2.7842776775360107,
|
| 1432 |
+
"eval_rewards/margins": 16.970603942871094,
|
| 1433 |
+
"eval_rewards/rejected": -19.754880905151367,
|
| 1434 |
+
"eval_runtime": 9.583,
|
| 1435 |
+
"eval_samples_per_second": 20.87,
|
| 1436 |
+
"eval_steps_per_second": 20.87,
|
| 1437 |
+
"step": 466
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 2.0085470085470085,
|
| 1441 |
+
"grad_norm": 0.080078125,
|
| 1442 |
+
"learning_rate": 6.974859870920561e-06,
|
| 1443 |
+
"logits/chosen": -1.3924753665924072,
|
| 1444 |
+
"logits/rejected": -0.8071160316467285,
|
| 1445 |
+
"logps/chosen": -68.68719482421875,
|
| 1446 |
+
"logps/rejected": -270.14801025390625,
|
| 1447 |
+
"loss": 0.0,
|
| 1448 |
+
"rewards/accuracies": 1.0,
|
| 1449 |
+
"rewards/chosen": -2.911695957183838,
|
| 1450 |
+
"rewards/margins": 16.461864471435547,
|
| 1451 |
+
"rewards/rejected": -19.37356185913086,
|
| 1452 |
+
"step": 470
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 2.02991452991453,
|
| 1456 |
+
"grad_norm": 0.0012359619140625,
|
| 1457 |
+
"learning_rate": 6.8548534975773135e-06,
|
| 1458 |
+
"logits/chosen": -1.4488104581832886,
|
| 1459 |
+
"logits/rejected": -0.82325679063797,
|
| 1460 |
+
"logps/chosen": -69.13402557373047,
|
| 1461 |
+
"logps/rejected": -275.86407470703125,
|
| 1462 |
+
"loss": 0.0001,
|
| 1463 |
+
"rewards/accuracies": 1.0,
|
| 1464 |
+
"rewards/chosen": -2.8428947925567627,
|
| 1465 |
+
"rewards/margins": 16.885896682739258,
|
| 1466 |
+
"rewards/rejected": -19.728792190551758,
|
| 1467 |
+
"step": 475
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 2.051282051282051,
|
| 1471 |
+
"grad_norm": 0.279296875,
|
| 1472 |
+
"learning_rate": 6.736370088322359e-06,
|
| 1473 |
+
"logits/chosen": -1.4575817584991455,
|
| 1474 |
+
"logits/rejected": -0.8154885172843933,
|
| 1475 |
+
"logps/chosen": -66.7756576538086,
|
| 1476 |
+
"logps/rejected": -272.9770812988281,
|
| 1477 |
+
"loss": 0.0006,
|
| 1478 |
+
"rewards/accuracies": 1.0,
|
| 1479 |
+
"rewards/chosen": -2.7416322231292725,
|
| 1480 |
+
"rewards/margins": 16.960206985473633,
|
| 1481 |
+
"rewards/rejected": -19.701839447021484,
|
| 1482 |
+
"step": 480
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 2.072649572649573,
|
| 1486 |
+
"grad_norm": 0.00994873046875,
|
| 1487 |
+
"learning_rate": 6.619475352261356e-06,
|
| 1488 |
+
"logits/chosen": -1.4201619625091553,
|
| 1489 |
+
"logits/rejected": -0.8027874827384949,
|
| 1490 |
+
"logps/chosen": -72.34085845947266,
|
| 1491 |
+
"logps/rejected": -276.4122009277344,
|
| 1492 |
+
"loss": 0.0,
|
| 1493 |
+
"rewards/accuracies": 1.0,
|
| 1494 |
+
"rewards/chosen": -3.0328619480133057,
|
| 1495 |
+
"rewards/margins": 16.768281936645508,
|
| 1496 |
+
"rewards/rejected": -19.801143646240234,
|
| 1497 |
+
"step": 485
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 2.094017094017094,
|
| 1501 |
+
"grad_norm": 0.0030975341796875,
|
| 1502 |
+
"learning_rate": 6.504234117445857e-06,
|
| 1503 |
+
"logits/chosen": -1.4316365718841553,
|
| 1504 |
+
"logits/rejected": -0.8082255125045776,
|
| 1505 |
+
"logps/chosen": -67.41413879394531,
|
| 1506 |
+
"logps/rejected": -272.8454284667969,
|
| 1507 |
+
"loss": 0.0001,
|
| 1508 |
+
"rewards/accuracies": 1.0,
|
| 1509 |
+
"rewards/chosen": -2.8370206356048584,
|
| 1510 |
+
"rewards/margins": 16.798355102539062,
|
| 1511 |
+
"rewards/rejected": -19.635372161865234,
|
| 1512 |
+
"step": 490
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 2.1153846153846154,
|
| 1516 |
+
"grad_norm": 0.012939453125,
|
| 1517 |
+
"learning_rate": 6.39071029492065e-06,
|
| 1518 |
+
"logits/chosen": -1.4187885522842407,
|
| 1519 |
+
"logits/rejected": -0.7696577906608582,
|
| 1520 |
+
"logps/chosen": -66.7591781616211,
|
| 1521 |
+
"logps/rejected": -271.50927734375,
|
| 1522 |
+
"loss": 0.0001,
|
| 1523 |
+
"rewards/accuracies": 1.0,
|
| 1524 |
+
"rewards/chosen": -2.810885190963745,
|
| 1525 |
+
"rewards/margins": 16.76914405822754,
|
| 1526 |
+
"rewards/rejected": -19.580028533935547,
|
| 1527 |
+
"step": 495
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 2.1367521367521367,
|
| 1531 |
+
"grad_norm": 0.125,
|
| 1532 |
+
"learning_rate": 6.2789668432796535e-06,
|
| 1533 |
+
"logits/chosen": -1.4300968647003174,
|
| 1534 |
+
"logits/rejected": -0.8352675437927246,
|
| 1535 |
+
"logps/chosen": -67.97314453125,
|
| 1536 |
+
"logps/rejected": -273.3833312988281,
|
| 1537 |
+
"loss": 0.0001,
|
| 1538 |
+
"rewards/accuracies": 1.0,
|
| 1539 |
+
"rewards/chosen": -2.8787360191345215,
|
| 1540 |
+
"rewards/margins": 16.89468002319336,
|
| 1541 |
+
"rewards/rejected": -19.773418426513672,
|
| 1542 |
+
"step": 500
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 2.1367521367521367,
|
| 1546 |
+
"eval_logits/chosen": -1.4944664239883423,
|
| 1547 |
+
"eval_logits/rejected": -0.8317099213600159,
|
| 1548 |
+
"eval_logps/chosen": -67.55956268310547,
|
| 1549 |
+
"eval_logps/rejected": -276.07452392578125,
|
| 1550 |
+
"eval_loss": 0.0001331747480435297,
|
| 1551 |
+
"eval_rewards/accuracies": 1.0,
|
| 1552 |
+
"eval_rewards/chosen": -2.811582565307617,
|
| 1553 |
+
"eval_rewards/margins": 16.994182586669922,
|
| 1554 |
+
"eval_rewards/rejected": -19.805763244628906,
|
| 1555 |
+
"eval_runtime": 9.5603,
|
| 1556 |
+
"eval_samples_per_second": 20.92,
|
| 1557 |
+
"eval_steps_per_second": 20.92,
|
| 1558 |
+
"step": 500
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 2.158119658119658,
|
| 1562 |
+
"grad_norm": 0.016845703125,
|
| 1563 |
+
"learning_rate": 6.16906573375004e-06,
|
| 1564 |
+
"logits/chosen": -1.4502363204956055,
|
| 1565 |
+
"logits/rejected": -0.8533867001533508,
|
| 1566 |
+
"logps/chosen": -67.79924774169922,
|
| 1567 |
+
"logps/rejected": -271.75946044921875,
|
| 1568 |
+
"loss": 0.0002,
|
| 1569 |
+
"rewards/accuracies": 1.0,
|
| 1570 |
+
"rewards/chosen": -2.8223788738250732,
|
| 1571 |
+
"rewards/margins": 16.64777374267578,
|
| 1572 |
+
"rewards/rejected": -19.47015380859375,
|
| 1573 |
+
"step": 505
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 2.1794871794871793,
|
| 1577 |
+
"grad_norm": 0.05908203125,
|
| 1578 |
+
"learning_rate": 6.061067915823923e-06,
|
| 1579 |
+
"logits/chosen": -1.3864041566848755,
|
| 1580 |
+
"logits/rejected": -0.7836092114448547,
|
| 1581 |
+
"logps/chosen": -65.47071838378906,
|
| 1582 |
+
"logps/rejected": -273.322265625,
|
| 1583 |
+
"loss": 0.0,
|
| 1584 |
+
"rewards/accuracies": 1.0,
|
| 1585 |
+
"rewards/chosen": -2.629401683807373,
|
| 1586 |
+
"rewards/margins": 16.939075469970703,
|
| 1587 |
+
"rewards/rejected": -19.568477630615234,
|
| 1588 |
+
"step": 510
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 2.200854700854701,
|
| 1592 |
+
"grad_norm": 0.006103515625,
|
| 1593 |
+
"learning_rate": 5.955033283456711e-06,
|
| 1594 |
+
"logits/chosen": -1.4336755275726318,
|
| 1595 |
+
"logits/rejected": -0.778075098991394,
|
| 1596 |
+
"logps/chosen": -70.93939971923828,
|
| 1597 |
+
"logps/rejected": -283.32598876953125,
|
| 1598 |
+
"loss": 0.0001,
|
| 1599 |
+
"rewards/accuracies": 1.0,
|
| 1600 |
+
"rewards/chosen": -3.026041030883789,
|
| 1601 |
+
"rewards/margins": 17.39080238342285,
|
| 1602 |
+
"rewards/rejected": -20.41684341430664,
|
| 1603 |
+
"step": 515
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 2.2222222222222223,
|
| 1607 |
+
"grad_norm": 0.00128936767578125,
|
| 1608 |
+
"learning_rate": 5.8510206418507914e-06,
|
| 1609 |
+
"logits/chosen": -1.4767441749572754,
|
| 1610 |
+
"logits/rejected": -0.8630167245864868,
|
| 1611 |
+
"logps/chosen": -68.52049255371094,
|
| 1612 |
+
"logps/rejected": -288.3775939941406,
|
| 1613 |
+
"loss": 0.0,
|
| 1614 |
+
"rewards/accuracies": 1.0,
|
| 1615 |
+
"rewards/chosen": -2.804711103439331,
|
| 1616 |
+
"rewards/margins": 18.02071762084961,
|
| 1617 |
+
"rewards/rejected": -20.825428009033203,
|
| 1618 |
+
"step": 520
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 2.2435897435897436,
|
| 1622 |
+
"grad_norm": 0.0078125,
|
| 1623 |
+
"learning_rate": 5.749087674843095e-06,
|
| 1624 |
+
"logits/chosen": -1.416854977607727,
|
| 1625 |
+
"logits/rejected": -0.8053888082504272,
|
| 1626 |
+
"logps/chosen": -63.85468673706055,
|
| 1627 |
+
"logps/rejected": -274.10955810546875,
|
| 1628 |
+
"loss": 0.0,
|
| 1629 |
+
"rewards/accuracies": 1.0,
|
| 1630 |
+
"rewards/chosen": -2.541356086730957,
|
| 1631 |
+
"rewards/margins": 17.136423110961914,
|
| 1632 |
+
"rewards/rejected": -19.67778205871582,
|
| 1633 |
+
"step": 525
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 2.264957264957265,
|
| 1637 |
+
"grad_norm": 0.024658203125,
|
| 1638 |
+
"learning_rate": 5.649290912914482e-06,
|
| 1639 |
+
"logits/chosen": -1.4132906198501587,
|
| 1640 |
+
"logits/rejected": -0.8062090873718262,
|
| 1641 |
+
"logps/chosen": -71.18649291992188,
|
| 1642 |
+
"logps/rejected": -287.32562255859375,
|
| 1643 |
+
"loss": 0.0,
|
| 1644 |
+
"rewards/accuracies": 1.0,
|
| 1645 |
+
"rewards/chosen": -2.9816391468048096,
|
| 1646 |
+
"rewards/margins": 17.66859245300293,
|
| 1647 |
+
"rewards/rejected": -20.650232315063477,
|
| 1648 |
+
"step": 530
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 2.286324786324786,
|
| 1652 |
+
"grad_norm": 0.031005859375,
|
| 1653 |
+
"learning_rate": 5.5516857018388144e-06,
|
| 1654 |
+
"logits/chosen": -1.4750016927719116,
|
| 1655 |
+
"logits/rejected": -0.8834052085876465,
|
| 1656 |
+
"logps/chosen": -66.85182189941406,
|
| 1657 |
+
"logps/rejected": -271.5949401855469,
|
| 1658 |
+
"loss": 0.0003,
|
| 1659 |
+
"rewards/accuracies": 1.0,
|
| 1660 |
+
"rewards/chosen": -2.7011666297912598,
|
| 1661 |
+
"rewards/margins": 16.737390518188477,
|
| 1662 |
+
"rewards/rejected": -19.438556671142578,
|
| 1663 |
+
"step": 535
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 2.3076923076923075,
|
| 1667 |
+
"grad_norm": 0.0098876953125,
|
| 1668 |
+
"learning_rate": 5.456326171989005e-06,
|
| 1669 |
+
"logits/chosen": -1.4350497722625732,
|
| 1670 |
+
"logits/rejected": -0.854056715965271,
|
| 1671 |
+
"logps/chosen": -64.83439636230469,
|
| 1672 |
+
"logps/rejected": -287.66961669921875,
|
| 1673 |
+
"loss": 0.0,
|
| 1674 |
+
"rewards/accuracies": 1.0,
|
| 1675 |
+
"rewards/chosen": -2.5173258781433105,
|
| 1676 |
+
"rewards/margins": 18.059844970703125,
|
| 1677 |
+
"rewards/rejected": -20.57716941833496,
|
| 1678 |
+
"step": 540
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 2.3290598290598292,
|
| 1682 |
+
"grad_norm": 0.0036163330078125,
|
| 1683 |
+
"learning_rate": 5.363265208317156e-06,
|
| 1684 |
+
"logits/chosen": -1.4079747200012207,
|
| 1685 |
+
"logits/rejected": -0.8292921781539917,
|
| 1686 |
+
"logps/chosen": -66.03783416748047,
|
| 1687 |
+
"logps/rejected": -271.29266357421875,
|
| 1688 |
+
"loss": 0.0,
|
| 1689 |
+
"rewards/accuracies": 1.0,
|
| 1690 |
+
"rewards/chosen": -2.6700127124786377,
|
| 1691 |
+
"rewards/margins": 16.861469268798828,
|
| 1692 |
+
"rewards/rejected": -19.531482696533203,
|
| 1693 |
+
"step": 545
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 2.3504273504273505,
|
| 1697 |
+
"grad_norm": 0.0035247802734375,
|
| 1698 |
+
"learning_rate": 5.272554421025347e-06,
|
| 1699 |
+
"logits/chosen": -1.4516395330429077,
|
| 1700 |
+
"logits/rejected": -0.8227602243423462,
|
| 1701 |
+
"logps/chosen": -68.09796142578125,
|
| 1702 |
+
"logps/rejected": -281.10687255859375,
|
| 1703 |
+
"loss": 0.0,
|
| 1704 |
+
"rewards/accuracies": 1.0,
|
| 1705 |
+
"rewards/chosen": -2.847126007080078,
|
| 1706 |
+
"rewards/margins": 17.43533706665039,
|
| 1707 |
+
"rewards/rejected": -20.2824649810791,
|
| 1708 |
+
"step": 550
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 2.371794871794872,
|
| 1712 |
+
"grad_norm": 0.04638671875,
|
| 1713 |
+
"learning_rate": 5.184244116943411e-06,
|
| 1714 |
+
"logits/chosen": -1.4369877576828003,
|
| 1715 |
+
"logits/rejected": -0.8126457333564758,
|
| 1716 |
+
"logps/chosen": -67.11319732666016,
|
| 1717 |
+
"logps/rejected": -276.0461730957031,
|
| 1718 |
+
"loss": 0.003,
|
| 1719 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 1720 |
+
"rewards/chosen": -2.69486665725708,
|
| 1721 |
+
"rewards/margins": 17.134822845458984,
|
| 1722 |
+
"rewards/rejected": -19.829692840576172,
|
| 1723 |
+
"step": 555
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"epoch": 2.393162393162393,
|
| 1727 |
+
"grad_norm": 0.013427734375,
|
| 1728 |
+
"learning_rate": 5.098383271629512e-06,
|
| 1729 |
+
"logits/chosen": -1.4552416801452637,
|
| 1730 |
+
"logits/rejected": -0.8165750503540039,
|
| 1731 |
+
"logps/chosen": -67.05858612060547,
|
| 1732 |
+
"logps/rejected": -270.05596923828125,
|
| 1733 |
+
"loss": 0.0001,
|
| 1734 |
+
"rewards/accuracies": 1.0,
|
| 1735 |
+
"rewards/chosen": -2.7700178623199463,
|
| 1736 |
+
"rewards/margins": 16.640972137451172,
|
| 1737 |
+
"rewards/rejected": -19.41098976135254,
|
| 1738 |
+
"step": 560
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 2.4145299145299144,
|
| 1742 |
+
"grad_norm": 0.024658203125,
|
| 1743 |
+
"learning_rate": 5.015019502209056e-06,
|
| 1744 |
+
"logits/chosen": -1.4252703189849854,
|
| 1745 |
+
"logits/rejected": -0.7998120188713074,
|
| 1746 |
+
"logps/chosen": -65.51549530029297,
|
| 1747 |
+
"logps/rejected": -266.0838623046875,
|
| 1748 |
+
"loss": 0.0,
|
| 1749 |
+
"rewards/accuracies": 1.0,
|
| 1750 |
+
"rewards/chosen": -2.6941094398498535,
|
| 1751 |
+
"rewards/margins": 16.32393455505371,
|
| 1752 |
+
"rewards/rejected": -19.018043518066406,
|
| 1753 |
+
"step": 565
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 2.435897435897436,
|
| 1757 |
+
"grad_norm": 0.00130462646484375,
|
| 1758 |
+
"learning_rate": 4.934199040966955e-06,
|
| 1759 |
+
"logits/chosen": -1.473314642906189,
|
| 1760 |
+
"logits/rejected": -0.847906231880188,
|
| 1761 |
+
"logps/chosen": -66.51988220214844,
|
| 1762 |
+
"logps/rejected": -269.98150634765625,
|
| 1763 |
+
"loss": 0.0001,
|
| 1764 |
+
"rewards/accuracies": 1.0,
|
| 1765 |
+
"rewards/chosen": -2.734851360321045,
|
| 1766 |
+
"rewards/margins": 16.677799224853516,
|
| 1767 |
+
"rewards/rejected": -19.412649154663086,
|
| 1768 |
+
"step": 570
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 2.4572649572649574,
|
| 1772 |
+
"grad_norm": 0.0120849609375,
|
| 1773 |
+
"learning_rate": 4.855966709707881e-06,
|
| 1774 |
+
"logits/chosen": -1.4144232273101807,
|
| 1775 |
+
"logits/rejected": -0.8331915736198425,
|
| 1776 |
+
"logps/chosen": -70.74285888671875,
|
| 1777 |
+
"logps/rejected": -275.6701354980469,
|
| 1778 |
+
"loss": 0.003,
|
| 1779 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 1780 |
+
"rewards/chosen": -3.0315637588500977,
|
| 1781 |
+
"rewards/margins": 16.809673309326172,
|
| 1782 |
+
"rewards/rejected": -19.841238021850586,
|
| 1783 |
+
"step": 575
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 2.4786324786324787,
|
| 1787 |
+
"grad_norm": 0.035888671875,
|
| 1788 |
+
"learning_rate": 4.780365894898799e-06,
|
| 1789 |
+
"logits/chosen": -1.446194052696228,
|
| 1790 |
+
"logits/rejected": -0.8259027600288391,
|
| 1791 |
+
"logps/chosen": -67.32369995117188,
|
| 1792 |
+
"logps/rejected": -276.65728759765625,
|
| 1793 |
+
"loss": 0.0,
|
| 1794 |
+
"rewards/accuracies": 1.0,
|
| 1795 |
+
"rewards/chosen": -2.730370044708252,
|
| 1796 |
+
"rewards/margins": 17.11151123046875,
|
| 1797 |
+
"rewards/rejected": -19.841882705688477,
|
| 1798 |
+
"step": 580
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 2.5,
|
| 1802 |
+
"grad_norm": 0.00885009765625,
|
| 1803 |
+
"learning_rate": 4.7074385236074684e-06,
|
| 1804 |
+
"logits/chosen": -1.4675248861312866,
|
| 1805 |
+
"logits/rejected": -0.8278427124023438,
|
| 1806 |
+
"logps/chosen": -71.28508758544922,
|
| 1807 |
+
"logps/rejected": -284.6138000488281,
|
| 1808 |
+
"loss": 0.0,
|
| 1809 |
+
"rewards/accuracies": 1.0,
|
| 1810 |
+
"rewards/chosen": -2.97928786277771,
|
| 1811 |
+
"rewards/margins": 17.529958724975586,
|
| 1812 |
+
"rewards/rejected": -20.509244918823242,
|
| 1813 |
+
"step": 585
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 2.5213675213675213,
|
| 1817 |
+
"grad_norm": 0.054931640625,
|
| 1818 |
+
"learning_rate": 4.63722504025034e-06,
|
| 1819 |
+
"logits/chosen": -1.4478528499603271,
|
| 1820 |
+
"logits/rejected": -0.8462594747543335,
|
| 1821 |
+
"logps/chosen": -64.45930480957031,
|
| 1822 |
+
"logps/rejected": -275.44036865234375,
|
| 1823 |
+
"loss": 0.0008,
|
| 1824 |
+
"rewards/accuracies": 1.0,
|
| 1825 |
+
"rewards/chosen": -2.4965498447418213,
|
| 1826 |
+
"rewards/margins": 17.28372573852539,
|
| 1827 |
+
"rewards/rejected": -19.780277252197266,
|
| 1828 |
+
"step": 590
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 2.5427350427350426,
|
| 1832 |
+
"grad_norm": 0.021240234375,
|
| 1833 |
+
"learning_rate": 4.569764384162676e-06,
|
| 1834 |
+
"logits/chosen": -1.4624818563461304,
|
| 1835 |
+
"logits/rejected": -0.8267906308174133,
|
| 1836 |
+
"logps/chosen": -61.47455596923828,
|
| 1837 |
+
"logps/rejected": -273.9830017089844,
|
| 1838 |
+
"loss": 0.0001,
|
| 1839 |
+
"rewards/accuracies": 1.0,
|
| 1840 |
+
"rewards/chosen": -2.310239791870117,
|
| 1841 |
+
"rewards/margins": 17.343975067138672,
|
| 1842 |
+
"rewards/rejected": -19.65421485900879,
|
| 1843 |
+
"step": 595
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 2.564102564102564,
|
| 1847 |
+
"grad_norm": 0.031494140625,
|
| 1848 |
+
"learning_rate": 4.50509396800341e-06,
|
| 1849 |
+
"logits/chosen": -1.4217302799224854,
|
| 1850 |
+
"logits/rejected": -0.7868741750717163,
|
| 1851 |
+
"logps/chosen": -65.87348937988281,
|
| 1852 |
+
"logps/rejected": -277.1817321777344,
|
| 1853 |
+
"loss": 0.0,
|
| 1854 |
+
"rewards/accuracies": 1.0,
|
| 1855 |
+
"rewards/chosen": -2.619393825531006,
|
| 1856 |
+
"rewards/margins": 17.432384490966797,
|
| 1857 |
+
"rewards/rejected": -20.051776885986328,
|
| 1858 |
+
"step": 600
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 2.5854700854700856,
|
| 1862 |
+
"grad_norm": 0.0732421875,
|
| 1863 |
+
"learning_rate": 4.443249657006627e-06,
|
| 1864 |
+
"logits/chosen": -1.4423940181732178,
|
| 1865 |
+
"logits/rejected": -0.7987694144248962,
|
| 1866 |
+
"logps/chosen": -63.83234786987305,
|
| 1867 |
+
"logps/rejected": -280.8373718261719,
|
| 1868 |
+
"loss": 0.0029,
|
| 1869 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 1870 |
+
"rewards/chosen": -2.4912962913513184,
|
| 1871 |
+
"rewards/margins": 17.850605010986328,
|
| 1872 |
+
"rewards/rejected": -20.341899871826172,
|
| 1873 |
+
"step": 605
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 2.606837606837607,
|
| 1877 |
+
"grad_norm": 0.01336669921875,
|
| 1878 |
+
"learning_rate": 4.384265749091266e-06,
|
| 1879 |
+
"logits/chosen": -1.391326665878296,
|
| 1880 |
+
"logits/rejected": -0.7768113613128662,
|
| 1881 |
+
"logps/chosen": -71.64138793945312,
|
| 1882 |
+
"logps/rejected": -278.91400146484375,
|
| 1883 |
+
"loss": 0.0001,
|
| 1884 |
+
"rewards/accuracies": 1.0,
|
| 1885 |
+
"rewards/chosen": -3.0774004459381104,
|
| 1886 |
+
"rewards/margins": 17.042530059814453,
|
| 1887 |
+
"rewards/rejected": -20.119930267333984,
|
| 1888 |
+
"step": 610
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 2.628205128205128,
|
| 1892 |
+
"grad_norm": 0.01239013671875,
|
| 1893 |
+
"learning_rate": 4.328174955840002e-06,
|
| 1894 |
+
"logits/chosen": -1.4435267448425293,
|
| 1895 |
+
"logits/rejected": -0.8103730082511902,
|
| 1896 |
+
"logps/chosen": -62.11974334716797,
|
| 1897 |
+
"logps/rejected": -273.298828125,
|
| 1898 |
+
"loss": 0.0,
|
| 1899 |
+
"rewards/accuracies": 1.0,
|
| 1900 |
+
"rewards/chosen": -2.468442440032959,
|
| 1901 |
+
"rewards/margins": 17.258960723876953,
|
| 1902 |
+
"rewards/rejected": -19.727405548095703,
|
| 1903 |
+
"step": 615
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 2.6495726495726495,
|
| 1907 |
+
"grad_norm": 0.004791259765625,
|
| 1908 |
+
"learning_rate": 4.275008384357902e-06,
|
| 1909 |
+
"logits/chosen": -1.4775636196136475,
|
| 1910 |
+
"logits/rejected": -0.8828374743461609,
|
| 1911 |
+
"logps/chosen": -65.2374267578125,
|
| 1912 |
+
"logps/rejected": -272.7725524902344,
|
| 1913 |
+
"loss": 0.0,
|
| 1914 |
+
"rewards/accuracies": 1.0,
|
| 1915 |
+
"rewards/chosen": -2.561352491378784,
|
| 1916 |
+
"rewards/margins": 17.067296981811523,
|
| 1917 |
+
"rewards/rejected": -19.62864875793457,
|
| 1918 |
+
"step": 620
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 2.6709401709401708,
|
| 1922 |
+
"grad_norm": 0.005157470703125,
|
| 1923 |
+
"learning_rate": 4.224795520020898e-06,
|
| 1924 |
+
"logits/chosen": -1.418306827545166,
|
| 1925 |
+
"logits/rejected": -0.7850183248519897,
|
| 1926 |
+
"logps/chosen": -68.51617431640625,
|
| 1927 |
+
"logps/rejected": -272.66455078125,
|
| 1928 |
+
"loss": 0.0001,
|
| 1929 |
+
"rewards/accuracies": 1.0,
|
| 1930 |
+
"rewards/chosen": -2.727930784225464,
|
| 1931 |
+
"rewards/margins": 16.756616592407227,
|
| 1932 |
+
"rewards/rejected": -19.484546661376953,
|
| 1933 |
+
"step": 625
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 2.6923076923076925,
|
| 1937 |
+
"grad_norm": 0.042724609375,
|
| 1938 |
+
"learning_rate": 4.177564210123634e-06,
|
| 1939 |
+
"logits/chosen": -1.4454832077026367,
|
| 1940 |
+
"logits/rejected": -0.8340616226196289,
|
| 1941 |
+
"logps/chosen": -65.76726531982422,
|
| 1942 |
+
"logps/rejected": -282.2267150878906,
|
| 1943 |
+
"loss": 0.0,
|
| 1944 |
+
"rewards/accuracies": 1.0,
|
| 1945 |
+
"rewards/chosen": -2.610382556915283,
|
| 1946 |
+
"rewards/margins": 17.712736129760742,
|
| 1947 |
+
"rewards/rejected": -20.323118209838867,
|
| 1948 |
+
"step": 630
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 2.713675213675214,
|
| 1952 |
+
"grad_norm": 0.039306640625,
|
| 1953 |
+
"learning_rate": 4.133340648435789e-06,
|
| 1954 |
+
"logits/chosen": -1.4548784494400024,
|
| 1955 |
+
"logits/rejected": -0.8171395063400269,
|
| 1956 |
+
"logps/chosen": -66.9704360961914,
|
| 1957 |
+
"logps/rejected": -277.67156982421875,
|
| 1958 |
+
"loss": 0.0001,
|
| 1959 |
+
"rewards/accuracies": 1.0,
|
| 1960 |
+
"rewards/chosen": -2.7575552463531494,
|
| 1961 |
+
"rewards/margins": 17.253223419189453,
|
| 1962 |
+
"rewards/rejected": -20.01078224182129,
|
| 1963 |
+
"step": 635
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 2.735042735042735,
|
| 1967 |
+
"grad_norm": 0.003448486328125,
|
| 1968 |
+
"learning_rate": 4.092149360675402e-06,
|
| 1969 |
+
"logits/chosen": -1.429237961769104,
|
| 1970 |
+
"logits/rejected": -0.825334906578064,
|
| 1971 |
+
"logps/chosen": -71.72950744628906,
|
| 1972 |
+
"logps/rejected": -282.79864501953125,
|
| 1973 |
+
"loss": 0.0,
|
| 1974 |
+
"rewards/accuracies": 1.0,
|
| 1975 |
+
"rewards/chosen": -2.9618561267852783,
|
| 1976 |
+
"rewards/margins": 17.245059967041016,
|
| 1977 |
+
"rewards/rejected": -20.206918716430664,
|
| 1978 |
+
"step": 640
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 2.7564102564102564,
|
| 1982 |
+
"grad_norm": 0.04833984375,
|
| 1983 |
+
"learning_rate": 4.054013190907282e-06,
|
| 1984 |
+
"logits/chosen": -1.3933217525482178,
|
| 1985 |
+
"logits/rejected": -0.7962290048599243,
|
| 1986 |
+
"logps/chosen": -62.87525177001953,
|
| 1987 |
+
"logps/rejected": -274.01885986328125,
|
| 1988 |
+
"loss": 0.0,
|
| 1989 |
+
"rewards/accuracies": 1.0,
|
| 1990 |
+
"rewards/chosen": -2.444106101989746,
|
| 1991 |
+
"rewards/margins": 17.38211441040039,
|
| 1992 |
+
"rewards/rejected": -19.826221466064453,
|
| 1993 |
+
"step": 645
|
| 1994 |
+
},
|
| 1995 |
+
{
|
| 1996 |
+
"epoch": 2.7777777777777777,
|
| 1997 |
+
"grad_norm": 0.0032196044921875,
|
| 1998 |
+
"learning_rate": 4.018953288874035e-06,
|
| 1999 |
+
"logits/chosen": -1.4579218626022339,
|
| 2000 |
+
"logits/rejected": -0.8607872724533081,
|
| 2001 |
+
"logps/chosen": -67.1677474975586,
|
| 2002 |
+
"logps/rejected": -278.09576416015625,
|
| 2003 |
+
"loss": 0.0,
|
| 2004 |
+
"rewards/accuracies": 1.0,
|
| 2005 |
+
"rewards/chosen": -2.7266063690185547,
|
| 2006 |
+
"rewards/margins": 17.331296920776367,
|
| 2007 |
+
"rewards/rejected": -20.057903289794922,
|
| 2008 |
+
"step": 650
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 2.799145299145299,
|
| 2012 |
+
"grad_norm": 0.0028533935546875,
|
| 2013 |
+
"learning_rate": 3.9869890982667385e-06,
|
| 2014 |
+
"logits/chosen": -1.4375386238098145,
|
| 2015 |
+
"logits/rejected": -0.7617470026016235,
|
| 2016 |
+
"logps/chosen": -66.19978332519531,
|
| 2017 |
+
"logps/rejected": -279.52581787109375,
|
| 2018 |
+
"loss": 0.0,
|
| 2019 |
+
"rewards/accuracies": 1.0,
|
| 2020 |
+
"rewards/chosen": -2.685457706451416,
|
| 2021 |
+
"rewards/margins": 17.297794342041016,
|
| 2022 |
+
"rewards/rejected": -19.983253479003906,
|
| 2023 |
+
"step": 655
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 2.8205128205128203,
|
| 2027 |
+
"grad_norm": 0.01300048828125,
|
| 2028 |
+
"learning_rate": 3.9581383459417625e-06,
|
| 2029 |
+
"logits/chosen": -1.409898042678833,
|
| 2030 |
+
"logits/rejected": -0.7993089556694031,
|
| 2031 |
+
"logps/chosen": -71.67328643798828,
|
| 2032 |
+
"logps/rejected": -286.96978759765625,
|
| 2033 |
+
"loss": 0.0,
|
| 2034 |
+
"rewards/accuracies": 1.0,
|
| 2035 |
+
"rewards/chosen": -3.136204242706299,
|
| 2036 |
+
"rewards/margins": 17.608535766601562,
|
| 2037 |
+
"rewards/rejected": -20.744739532470703,
|
| 2038 |
+
"step": 660
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 2.841880341880342,
|
| 2042 |
+
"grad_norm": 0.029052734375,
|
| 2043 |
+
"learning_rate": 3.932417032089722e-06,
|
| 2044 |
+
"logits/chosen": -1.4576003551483154,
|
| 2045 |
+
"logits/rejected": -0.821110725402832,
|
| 2046 |
+
"logps/chosen": -68.1684799194336,
|
| 2047 |
+
"logps/rejected": -282.56744384765625,
|
| 2048 |
+
"loss": 0.0001,
|
| 2049 |
+
"rewards/accuracies": 1.0,
|
| 2050 |
+
"rewards/chosen": -2.7884955406188965,
|
| 2051 |
+
"rewards/margins": 17.563434600830078,
|
| 2052 |
+
"rewards/rejected": -20.351930618286133,
|
| 2053 |
+
"step": 665
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 2.8632478632478633,
|
| 2057 |
+
"grad_norm": 0.000885009765625,
|
| 2058 |
+
"learning_rate": 3.909839421362017e-06,
|
| 2059 |
+
"logits/chosen": -1.4005993604660034,
|
| 2060 |
+
"logits/rejected": -0.7887626886367798,
|
| 2061 |
+
"logps/chosen": -67.8237075805664,
|
| 2062 |
+
"logps/rejected": -278.14056396484375,
|
| 2063 |
+
"loss": 0.0001,
|
| 2064 |
+
"rewards/accuracies": 1.0,
|
| 2065 |
+
"rewards/chosen": -2.829472064971924,
|
| 2066 |
+
"rewards/margins": 17.069793701171875,
|
| 2067 |
+
"rewards/rejected": -19.899267196655273,
|
| 2068 |
+
"step": 670
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 2.8846153846153846,
|
| 2072 |
+
"grad_norm": 0.048828125,
|
| 2073 |
+
"learning_rate": 3.890418034959871e-06,
|
| 2074 |
+
"logits/chosen": -1.3995271921157837,
|
| 2075 |
+
"logits/rejected": -0.7455986738204956,
|
| 2076 |
+
"logps/chosen": -66.70311737060547,
|
| 2077 |
+
"logps/rejected": -276.72540283203125,
|
| 2078 |
+
"loss": 0.0,
|
| 2079 |
+
"rewards/accuracies": 1.0,
|
| 2080 |
+
"rewards/chosen": -2.7374672889709473,
|
| 2081 |
+
"rewards/margins": 17.271923065185547,
|
| 2082 |
+
"rewards/rejected": -20.00939178466797,
|
| 2083 |
+
"step": 675
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 2.905982905982906,
|
| 2087 |
+
"grad_norm": 0.045166015625,
|
| 2088 |
+
"learning_rate": 3.874163643690263e-06,
|
| 2089 |
+
"logits/chosen": -1.3939011096954346,
|
| 2090 |
+
"logits/rejected": -0.7581735849380493,
|
| 2091 |
+
"logps/chosen": -74.34012603759766,
|
| 2092 |
+
"logps/rejected": -280.92291259765625,
|
| 2093 |
+
"loss": 0.0,
|
| 2094 |
+
"rewards/accuracies": 1.0,
|
| 2095 |
+
"rewards/chosen": -3.3185927867889404,
|
| 2096 |
+
"rewards/margins": 16.952451705932617,
|
| 2097 |
+
"rewards/rejected": -20.27104377746582,
|
| 2098 |
+
"step": 680
|
| 2099 |
+
},
|
| 2100 |
+
{
|
| 2101 |
+
"epoch": 2.9273504273504276,
|
| 2102 |
+
"grad_norm": 0.01312255859375,
|
| 2103 |
+
"learning_rate": 3.861085261992599e-06,
|
| 2104 |
+
"logits/chosen": -1.4342122077941895,
|
| 2105 |
+
"logits/rejected": -0.8670775294303894,
|
| 2106 |
+
"logps/chosen": -70.84014129638672,
|
| 2107 |
+
"logps/rejected": -280.0901794433594,
|
| 2108 |
+
"loss": 0.0,
|
| 2109 |
+
"rewards/accuracies": 1.0,
|
| 2110 |
+
"rewards/chosen": -3.0605998039245605,
|
| 2111 |
+
"rewards/margins": 17.057954788208008,
|
| 2112 |
+
"rewards/rejected": -20.118555068969727,
|
| 2113 |
+
"step": 685
|
| 2114 |
+
},
|
| 2115 |
+
{
|
| 2116 |
+
"epoch": 2.948717948717949,
|
| 2117 |
+
"grad_norm": 0.1376953125,
|
| 2118 |
+
"learning_rate": 3.851190142939442e-06,
|
| 2119 |
+
"logits/chosen": -1.435960054397583,
|
| 2120 |
+
"logits/rejected": -0.8204304575920105,
|
| 2121 |
+
"logps/chosen": -65.02848052978516,
|
| 2122 |
+
"logps/rejected": -277.8736877441406,
|
| 2123 |
+
"loss": 0.0001,
|
| 2124 |
+
"rewards/accuracies": 1.0,
|
| 2125 |
+
"rewards/chosen": -2.4977402687072754,
|
| 2126 |
+
"rewards/margins": 17.434282302856445,
|
| 2127 |
+
"rewards/rejected": -19.932022094726562,
|
| 2128 |
+
"step": 690
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"epoch": 2.97008547008547,
|
| 2132 |
+
"grad_norm": 0.003662109375,
|
| 2133 |
+
"learning_rate": 3.844483774214069e-06,
|
| 2134 |
+
"logits/chosen": -1.4186488389968872,
|
| 2135 |
+
"logits/rejected": -0.7549436688423157,
|
| 2136 |
+
"logps/chosen": -64.54364776611328,
|
| 2137 |
+
"logps/rejected": -277.66680908203125,
|
| 2138 |
+
"loss": 0.0002,
|
| 2139 |
+
"rewards/accuracies": 1.0,
|
| 2140 |
+
"rewards/chosen": -2.5480058193206787,
|
| 2141 |
+
"rewards/margins": 17.503028869628906,
|
| 2142 |
+
"rewards/rejected": -20.0510311126709,
|
| 2143 |
+
"step": 695
|
| 2144 |
+
},
|
| 2145 |
+
{
|
| 2146 |
+
"epoch": 2.9871794871794872,
|
| 2147 |
+
"eval_logits/chosen": -1.4855865240097046,
|
| 2148 |
+
"eval_logits/rejected": -0.8208981156349182,
|
| 2149 |
+
"eval_logps/chosen": -67.3394775390625,
|
| 2150 |
+
"eval_logps/rejected": -276.6026306152344,
|
| 2151 |
+
"eval_loss": 0.00011922841804334894,
|
| 2152 |
+
"eval_rewards/accuracies": 1.0,
|
| 2153 |
+
"eval_rewards/chosen": -2.789573907852173,
|
| 2154 |
+
"eval_rewards/margins": 17.068998336791992,
|
| 2155 |
+
"eval_rewards/rejected": -19.858572006225586,
|
| 2156 |
+
"eval_runtime": 9.6169,
|
| 2157 |
+
"eval_samples_per_second": 20.797,
|
| 2158 |
+
"eval_steps_per_second": 20.797,
|
| 2159 |
+
"step": 699
|
| 2160 |
+
}
|
| 2161 |
+
],
|
| 2162 |
+
"logging_steps": 5,
|
| 2163 |
+
"max_steps": 702,
|
| 2164 |
+
"num_input_tokens_seen": 0,
|
| 2165 |
+
"num_train_epochs": 3,
|
| 2166 |
+
"save_steps": 500,
|
| 2167 |
+
"stateful_callbacks": {
|
| 2168 |
+
"TrainerControl": {
|
| 2169 |
+
"args": {
|
| 2170 |
+
"should_epoch_stop": false,
|
| 2171 |
+
"should_evaluate": false,
|
| 2172 |
+
"should_log": false,
|
| 2173 |
+
"should_save": true,
|
| 2174 |
+
"should_training_stop": false
|
| 2175 |
+
},
|
| 2176 |
+
"attributes": {}
|
| 2177 |
+
}
|
| 2178 |
+
},
|
| 2179 |
+
"total_flos": 0.0,
|
| 2180 |
+
"train_batch_size": 12,
|
| 2181 |
+
"trial_name": null,
|
| 2182 |
+
"trial_params": null
|
| 2183 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ff0614c02dc99ec0582da6d35bb87b67a526fd0a58658c483d24d2bfe5db1412
|
| 3 |
+
size 6520
|