sravanthib commited on
Commit
410e7b9
·
verified ·
1 Parent(s): 36e6715

Training in progress, step 10

Browse files
README.md CHANGED
@@ -3,19 +3,16 @@ library_name: peft
3
  license: llama3.2
4
  base_model: meta-llama/Llama-3.2-1B
5
  tags:
6
- - base_model:adapter:meta-llama/Llama-3.2-1B
7
- - lora
8
- - transformers
9
- pipeline_tag: text-generation
10
  model-index:
11
- - name: testing
12
  results: []
13
  ---
14
 
15
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
16
  should probably proofread and complete it, then remove this comment. -->
17
 
18
- # testing
19
 
20
  This model is a fine-tuned version of [meta-llama/Llama-3.2-1B](https://huggingface.co/meta-llama/Llama-3.2-1B) on an unknown dataset.
21
 
@@ -45,10 +42,10 @@ The following hyperparameters were used during training:
45
  - gradient_accumulation_steps: 10
46
  - total_train_batch_size: 160
47
  - total_eval_batch_size: 64
48
- - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
49
  - lr_scheduler_type: cosine
50
- - lr_scheduler_warmup_ratio: 0.03
51
- - training_steps: 20
52
 
53
  ### Training results
54
 
@@ -56,8 +53,8 @@ The following hyperparameters were used during training:
56
 
57
  ### Framework versions
58
 
59
- - PEFT 0.16.0
60
- - Transformers 4.53.2
61
- - Pytorch 2.7.1+cu126
62
- - Datasets 4.0.0
63
  - Tokenizers 0.21.2
 
3
  license: llama3.2
4
  base_model: meta-llama/Llama-3.2-1B
5
  tags:
6
+ - generated_from_trainer
 
 
 
7
  model-index:
8
+ - name: multi-gpu-llama-3-2-1b-40k-1e-4-custom-sft-2048-seqlen
9
  results: []
10
  ---
11
 
12
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
13
  should probably proofread and complete it, then remove this comment. -->
14
 
15
+ # multi-gpu-llama-3-2-1b-40k-1e-4-custom-sft-2048-seqlen
16
 
17
  This model is a fine-tuned version of [meta-llama/Llama-3.2-1B](https://huggingface.co/meta-llama/Llama-3.2-1B) on an unknown dataset.
18
 
 
42
  - gradient_accumulation_steps: 10
43
  - total_train_batch_size: 160
44
  - total_eval_batch_size: 64
45
+ - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
46
  - lr_scheduler_type: cosine
47
+ - lr_scheduler_warmup_ratio: 0.05
48
+ - training_steps: 1000
49
 
50
  ### Training results
51
 
 
53
 
54
  ### Framework versions
55
 
56
+ - PEFT 0.14.0
57
+ - Transformers 4.51.3
58
+ - Pytorch 2.3.0+cu121
59
+ - Datasets 2.15.0
60
  - Tokenizers 0.21.2
adapter_config.json CHANGED
@@ -3,7 +3,6 @@
3
  "auto_mapping": null,
4
  "base_model_name_or_path": "meta-llama/Llama-3.2-1B",
5
  "bias": "none",
6
- "corda_config": null,
7
  "eva_config": null,
8
  "exclude_modules": null,
9
  "fan_in_fan_out": false,
@@ -13,29 +12,26 @@
13
  "layers_pattern": null,
14
  "layers_to_transform": null,
15
  "loftq_config": {},
16
- "lora_alpha": 16,
17
  "lora_bias": false,
18
  "lora_dropout": 0.05,
19
  "megatron_config": null,
20
  "megatron_core": "megatron.core",
21
  "modules_to_save": null,
22
  "peft_type": "LORA",
23
- "qalora_group_size": 16,
24
  "r": 8,
25
  "rank_pattern": {},
26
  "revision": null,
27
  "target_modules": [
28
- "k_proj",
29
  "v_proj",
 
30
  "o_proj",
31
- "down_proj",
32
  "q_proj",
33
- "gate_proj",
34
- "up_proj"
35
  ],
36
  "task_type": "CAUSAL_LM",
37
- "trainable_token_indices": null,
38
  "use_dora": false,
39
- "use_qalora": false,
40
  "use_rslora": false
41
  }
 
3
  "auto_mapping": null,
4
  "base_model_name_or_path": "meta-llama/Llama-3.2-1B",
5
  "bias": "none",
 
6
  "eva_config": null,
7
  "exclude_modules": null,
8
  "fan_in_fan_out": false,
 
12
  "layers_pattern": null,
13
  "layers_to_transform": null,
14
  "loftq_config": {},
15
+ "lora_alpha": 16.0,
16
  "lora_bias": false,
17
  "lora_dropout": 0.05,
18
  "megatron_config": null,
19
  "megatron_core": "megatron.core",
20
  "modules_to_save": null,
21
  "peft_type": "LORA",
 
22
  "r": 8,
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
26
+ "gate_proj",
27
  "v_proj",
28
+ "up_proj",
29
  "o_proj",
 
30
  "q_proj",
31
+ "down_proj",
32
+ "k_proj"
33
  ],
34
  "task_type": "CAUSAL_LM",
 
35
  "use_dora": false,
 
36
  "use_rslora": false
37
  }
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b9a4110b2f067c1c3df33d9875e16ff46ec4326efa78cd8ecf7e6b17dfa169e1
3
  size 11301520
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:daf11edf266a10e84e0736bf55479df8aadd34e14d6515e4f3e2b5c89e4ebecc
3
  size 11301520
added_tokens.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</tool_call>": 151658,
3
+ "<tool_call>": 151657,
4
+ "<|box_end|>": 151649,
5
+ "<|box_start|>": 151648,
6
+ "<|endoftext|>": 151643,
7
+ "<|file_sep|>": 151664,
8
+ "<|fim_middle|>": 151660,
9
+ "<|fim_pad|>": 151662,
10
+ "<|fim_prefix|>": 151659,
11
+ "<|fim_suffix|>": 151661,
12
+ "<|im_end|>": 151645,
13
+ "<|im_start|>": 151644,
14
+ "<|image_pad|>": 151655,
15
+ "<|object_ref_end|>": 151647,
16
+ "<|object_ref_start|>": 151646,
17
+ "<|quad_end|>": 151651,
18
+ "<|quad_start|>": 151650,
19
+ "<|repo_name|>": 151663,
20
+ "<|video_pad|>": 151656,
21
+ "<|vision_end|>": 151653,
22
+ "<|vision_pad|>": 151654,
23
+ "<|vision_start|>": 151652
24
+ }
all_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "epoch": 0.0365296803652968,
3
- "total_flos": 3.848728354383462e+16,
4
- "train_loss": 1.5816513061523438,
5
- "train_runtime": 267.2013,
6
- "train_samples_per_second": 11.976,
7
- "train_steps_per_second": 0.075
8
  }
 
1
  {
2
+ "epoch": 4.0,
3
+ "total_flos": 1.9243641771917312e+18,
4
+ "train_loss": 0.050279883230105044,
5
+ "train_runtime": 7527.0065,
6
+ "train_samples_per_second": 21.257,
7
+ "train_steps_per_second": 0.133
8
  }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json CHANGED
@@ -2060,5 +2060,5 @@
2060
  "model_max_length": 131072,
2061
  "pad_token": "<|end_of_text|>",
2062
  "padding_side": "right",
2063
- "tokenizer_class": "PreTrainedTokenizerFast"
2064
  }
 
2060
  "model_max_length": 131072,
2061
  "pad_token": "<|end_of_text|>",
2062
  "padding_side": "right",
2063
+ "tokenizer_class": "PreTrainedTokenizer"
2064
  }
train_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "epoch": 0.0365296803652968,
3
- "total_flos": 3.848728354383462e+16,
4
- "train_loss": 1.5816513061523438,
5
- "train_runtime": 267.2013,
6
- "train_samples_per_second": 11.976,
7
- "train_steps_per_second": 0.075
8
  }
 
1
  {
2
+ "epoch": 4.0,
3
+ "total_flos": 1.9243641771917312e+18,
4
+ "train_loss": 0.050279883230105044,
5
+ "train_runtime": 7527.0065,
6
+ "train_samples_per_second": 21.257,
7
+ "train_steps_per_second": 0.133
8
  }
trainer_state.json CHANGED
@@ -2,27 +2,727 @@
2
  "best_global_step": null,
3
  "best_metric": null,
4
  "best_model_checkpoint": null,
5
- "epoch": 0.0365296803652968,
6
  "eval_steps": 0,
7
- "global_step": 20,
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
11
  "log_history": [
12
  {
13
- "epoch": 0.0365296803652968,
14
- "step": 20,
15
- "total_flos": 3.848728354383462e+16,
16
- "train_loss": 1.5816513061523438,
17
- "train_runtime": 267.2013,
18
- "train_samples_per_second": 11.976,
19
- "train_steps_per_second": 0.075
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20
  }
21
  ],
22
- "logging_steps": 100,
23
- "max_steps": 20,
24
  "num_input_tokens_seen": 0,
25
- "num_train_epochs": 1,
26
  "save_steps": 500,
27
  "stateful_callbacks": {
28
  "TrainerControl": {
@@ -36,7 +736,7 @@
36
  "attributes": {}
37
  }
38
  },
39
- "total_flos": 3.848728354383462e+16,
40
  "train_batch_size": 2,
41
  "trial_name": null,
42
  "trial_params": null
 
2
  "best_global_step": null,
3
  "best_metric": null,
4
  "best_model_checkpoint": null,
5
+ "epoch": 4.0,
6
  "eval_steps": 0,
7
+ "global_step": 1000,
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
11
  "log_history": [
12
  {
13
+ "epoch": 0.04,
14
+ "grad_norm": 2.592644453048706,
15
+ "learning_rate": 5.8859191006777906e-05,
16
+ "loss": 4.3679,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.08,
21
+ "grad_norm": 0.18703584372997284,
22
+ "learning_rate": 7.657757302033369e-05,
23
+ "loss": 0.0344,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.12,
28
+ "grad_norm": 0.10068075358867645,
29
+ "learning_rate": 8.694216207171606e-05,
30
+ "loss": 0.0263,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.16,
35
+ "grad_norm": 0.06798414140939713,
36
+ "learning_rate": 9.429595503388949e-05,
37
+ "loss": 0.0251,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.2,
42
+ "grad_norm": 0.0597367137670517,
43
+ "learning_rate": 0.0001,
44
+ "loss": 0.0245,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.24,
49
+ "grad_norm": 0.04497462511062622,
50
+ "learning_rate": 0.0001,
51
+ "loss": 0.0236,
52
+ "step": 60
53
+ },
54
+ {
55
+ "epoch": 0.28,
56
+ "grad_norm": 0.04295974224805832,
57
+ "learning_rate": 0.0001,
58
+ "loss": 0.0231,
59
+ "step": 70
60
+ },
61
+ {
62
+ "epoch": 0.32,
63
+ "grad_norm": 0.0416204072535038,
64
+ "learning_rate": 0.0001,
65
+ "loss": 0.0225,
66
+ "step": 80
67
+ },
68
+ {
69
+ "epoch": 0.36,
70
+ "grad_norm": 0.03994525969028473,
71
+ "learning_rate": 0.0001,
72
+ "loss": 0.022,
73
+ "step": 90
74
+ },
75
+ {
76
+ "epoch": 0.4,
77
+ "grad_norm": 0.045212067663669586,
78
+ "learning_rate": 0.0001,
79
+ "loss": 0.0215,
80
+ "step": 100
81
+ },
82
+ {
83
+ "epoch": 0.44,
84
+ "grad_norm": 0.034988779574632645,
85
+ "learning_rate": 0.0001,
86
+ "loss": 0.0209,
87
+ "step": 110
88
+ },
89
+ {
90
+ "epoch": 0.48,
91
+ "grad_norm": 0.03513108938932419,
92
+ "learning_rate": 0.0001,
93
+ "loss": 0.0205,
94
+ "step": 120
95
+ },
96
+ {
97
+ "epoch": 0.52,
98
+ "grad_norm": 0.03687143325805664,
99
+ "learning_rate": 0.0001,
100
+ "loss": 0.0197,
101
+ "step": 130
102
+ },
103
+ {
104
+ "epoch": 0.56,
105
+ "grad_norm": 0.031846143305301666,
106
+ "learning_rate": 0.0001,
107
+ "loss": 0.0192,
108
+ "step": 140
109
+ },
110
+ {
111
+ "epoch": 0.6,
112
+ "grad_norm": 0.04322415590286255,
113
+ "learning_rate": 0.0001,
114
+ "loss": 0.0184,
115
+ "step": 150
116
+ },
117
+ {
118
+ "epoch": 0.64,
119
+ "grad_norm": 0.03877222165465355,
120
+ "learning_rate": 0.0001,
121
+ "loss": 0.0176,
122
+ "step": 160
123
+ },
124
+ {
125
+ "epoch": 0.68,
126
+ "grad_norm": 0.03259311988949776,
127
+ "learning_rate": 0.0001,
128
+ "loss": 0.0165,
129
+ "step": 170
130
+ },
131
+ {
132
+ "epoch": 0.72,
133
+ "grad_norm": 0.04251185432076454,
134
+ "learning_rate": 0.0001,
135
+ "loss": 0.0151,
136
+ "step": 180
137
+ },
138
+ {
139
+ "epoch": 0.76,
140
+ "grad_norm": 0.055991314351558685,
141
+ "learning_rate": 0.0001,
142
+ "loss": 0.0135,
143
+ "step": 190
144
+ },
145
+ {
146
+ "epoch": 0.8,
147
+ "grad_norm": 0.049590013921260834,
148
+ "learning_rate": 0.0001,
149
+ "loss": 0.0112,
150
+ "step": 200
151
+ },
152
+ {
153
+ "epoch": 0.84,
154
+ "grad_norm": 0.029243191704154015,
155
+ "learning_rate": 0.0001,
156
+ "loss": 0.0097,
157
+ "step": 210
158
+ },
159
+ {
160
+ "epoch": 0.88,
161
+ "grad_norm": 0.028539329767227173,
162
+ "learning_rate": 0.0001,
163
+ "loss": 0.0087,
164
+ "step": 220
165
+ },
166
+ {
167
+ "epoch": 0.92,
168
+ "grad_norm": 0.02026432566344738,
169
+ "learning_rate": 0.0001,
170
+ "loss": 0.008,
171
+ "step": 230
172
+ },
173
+ {
174
+ "epoch": 0.96,
175
+ "grad_norm": 0.017264919355511665,
176
+ "learning_rate": 0.0001,
177
+ "loss": 0.0076,
178
+ "step": 240
179
+ },
180
+ {
181
+ "epoch": 1.0,
182
+ "grad_norm": 0.022517086938023567,
183
+ "learning_rate": 0.0001,
184
+ "loss": 0.0071,
185
+ "step": 250
186
+ },
187
+ {
188
+ "epoch": 1.04,
189
+ "grad_norm": 0.016512203961610794,
190
+ "learning_rate": 0.0001,
191
+ "loss": 0.007,
192
+ "step": 260
193
+ },
194
+ {
195
+ "epoch": 1.08,
196
+ "grad_norm": 0.01752372272312641,
197
+ "learning_rate": 0.0001,
198
+ "loss": 0.0065,
199
+ "step": 270
200
+ },
201
+ {
202
+ "epoch": 1.12,
203
+ "grad_norm": 0.014535382390022278,
204
+ "learning_rate": 0.0001,
205
+ "loss": 0.006,
206
+ "step": 280
207
+ },
208
+ {
209
+ "epoch": 1.16,
210
+ "grad_norm": 0.01353287324309349,
211
+ "learning_rate": 0.0001,
212
+ "loss": 0.0057,
213
+ "step": 290
214
+ },
215
+ {
216
+ "epoch": 1.2,
217
+ "grad_norm": 0.014846131205558777,
218
+ "learning_rate": 0.0001,
219
+ "loss": 0.0055,
220
+ "step": 300
221
+ },
222
+ {
223
+ "epoch": 1.24,
224
+ "grad_norm": 0.01546796876937151,
225
+ "learning_rate": 0.0001,
226
+ "loss": 0.0052,
227
+ "step": 310
228
+ },
229
+ {
230
+ "epoch": 1.28,
231
+ "grad_norm": 0.013868402689695358,
232
+ "learning_rate": 0.0001,
233
+ "loss": 0.0049,
234
+ "step": 320
235
+ },
236
+ {
237
+ "epoch": 1.32,
238
+ "grad_norm": 0.013835090212523937,
239
+ "learning_rate": 0.0001,
240
+ "loss": 0.0049,
241
+ "step": 330
242
+ },
243
+ {
244
+ "epoch": 1.3599999999999999,
245
+ "grad_norm": 0.01379898190498352,
246
+ "learning_rate": 0.0001,
247
+ "loss": 0.0047,
248
+ "step": 340
249
+ },
250
+ {
251
+ "epoch": 1.4,
252
+ "grad_norm": 0.014727466739714146,
253
+ "learning_rate": 0.0001,
254
+ "loss": 0.0045,
255
+ "step": 350
256
+ },
257
+ {
258
+ "epoch": 1.44,
259
+ "grad_norm": 0.011309986934065819,
260
+ "learning_rate": 0.0001,
261
+ "loss": 0.0044,
262
+ "step": 360
263
+ },
264
+ {
265
+ "epoch": 1.48,
266
+ "grad_norm": 0.021115651354193687,
267
+ "learning_rate": 0.0001,
268
+ "loss": 0.0042,
269
+ "step": 370
270
+ },
271
+ {
272
+ "epoch": 1.52,
273
+ "grad_norm": 0.01117760967463255,
274
+ "learning_rate": 0.0001,
275
+ "loss": 0.0041,
276
+ "step": 380
277
+ },
278
+ {
279
+ "epoch": 1.56,
280
+ "grad_norm": 0.010027006268501282,
281
+ "learning_rate": 0.0001,
282
+ "loss": 0.0041,
283
+ "step": 390
284
+ },
285
+ {
286
+ "epoch": 1.6,
287
+ "grad_norm": 0.012770233675837517,
288
+ "learning_rate": 0.0001,
289
+ "loss": 0.0039,
290
+ "step": 400
291
+ },
292
+ {
293
+ "epoch": 1.6400000000000001,
294
+ "grad_norm": 0.012774297036230564,
295
+ "learning_rate": 0.0001,
296
+ "loss": 0.0037,
297
+ "step": 410
298
+ },
299
+ {
300
+ "epoch": 1.6800000000000002,
301
+ "grad_norm": 0.018494736403226852,
302
+ "learning_rate": 0.0001,
303
+ "loss": 0.0038,
304
+ "step": 420
305
+ },
306
+ {
307
+ "epoch": 1.72,
308
+ "grad_norm": 0.010668110102415085,
309
+ "learning_rate": 0.0001,
310
+ "loss": 0.0035,
311
+ "step": 430
312
+ },
313
+ {
314
+ "epoch": 1.76,
315
+ "grad_norm": 0.009887272492051125,
316
+ "learning_rate": 0.0001,
317
+ "loss": 0.0036,
318
+ "step": 440
319
+ },
320
+ {
321
+ "epoch": 1.8,
322
+ "grad_norm": 0.009353808127343655,
323
+ "learning_rate": 0.0001,
324
+ "loss": 0.0034,
325
+ "step": 450
326
+ },
327
+ {
328
+ "epoch": 1.8399999999999999,
329
+ "grad_norm": 0.009478743188083172,
330
+ "learning_rate": 0.0001,
331
+ "loss": 0.0034,
332
+ "step": 460
333
+ },
334
+ {
335
+ "epoch": 1.88,
336
+ "grad_norm": 0.012139740400016308,
337
+ "learning_rate": 0.0001,
338
+ "loss": 0.0034,
339
+ "step": 470
340
+ },
341
+ {
342
+ "epoch": 1.92,
343
+ "grad_norm": 0.011693431995809078,
344
+ "learning_rate": 0.0001,
345
+ "loss": 0.0034,
346
+ "step": 480
347
+ },
348
+ {
349
+ "epoch": 1.96,
350
+ "grad_norm": 0.009398797526955605,
351
+ "learning_rate": 0.0001,
352
+ "loss": 0.0031,
353
+ "step": 490
354
+ },
355
+ {
356
+ "epoch": 2.0,
357
+ "grad_norm": 0.008883015252649784,
358
+ "learning_rate": 0.0001,
359
+ "loss": 0.0032,
360
+ "step": 500
361
+ },
362
+ {
363
+ "epoch": 2.04,
364
+ "grad_norm": 0.00975903868675232,
365
+ "learning_rate": 0.0001,
366
+ "loss": 0.0031,
367
+ "step": 510
368
+ },
369
+ {
370
+ "epoch": 2.08,
371
+ "grad_norm": 0.009527013637125492,
372
+ "learning_rate": 0.0001,
373
+ "loss": 0.0029,
374
+ "step": 520
375
+ },
376
+ {
377
+ "epoch": 2.12,
378
+ "grad_norm": 0.011014975607395172,
379
+ "learning_rate": 0.0001,
380
+ "loss": 0.003,
381
+ "step": 530
382
+ },
383
+ {
384
+ "epoch": 2.16,
385
+ "grad_norm": 0.00897703692317009,
386
+ "learning_rate": 0.0001,
387
+ "loss": 0.0029,
388
+ "step": 540
389
+ },
390
+ {
391
+ "epoch": 2.2,
392
+ "grad_norm": 0.009742386639118195,
393
+ "learning_rate": 0.0001,
394
+ "loss": 0.0028,
395
+ "step": 550
396
+ },
397
+ {
398
+ "epoch": 2.24,
399
+ "grad_norm": 0.010376248508691788,
400
+ "learning_rate": 0.0001,
401
+ "loss": 0.0028,
402
+ "step": 560
403
+ },
404
+ {
405
+ "epoch": 2.2800000000000002,
406
+ "grad_norm": 0.010401038452982903,
407
+ "learning_rate": 0.0001,
408
+ "loss": 0.0028,
409
+ "step": 570
410
+ },
411
+ {
412
+ "epoch": 2.32,
413
+ "grad_norm": 0.009547604247927666,
414
+ "learning_rate": 0.0001,
415
+ "loss": 0.0027,
416
+ "step": 580
417
+ },
418
+ {
419
+ "epoch": 2.36,
420
+ "grad_norm": 0.007607370615005493,
421
+ "learning_rate": 0.0001,
422
+ "loss": 0.0026,
423
+ "step": 590
424
+ },
425
+ {
426
+ "epoch": 2.4,
427
+ "grad_norm": 0.008813534863293171,
428
+ "learning_rate": 0.0001,
429
+ "loss": 0.0026,
430
+ "step": 600
431
+ },
432
+ {
433
+ "epoch": 2.44,
434
+ "grad_norm": 0.010278712958097458,
435
+ "learning_rate": 0.0001,
436
+ "loss": 0.0026,
437
+ "step": 610
438
+ },
439
+ {
440
+ "epoch": 2.48,
441
+ "grad_norm": 0.011279788799583912,
442
+ "learning_rate": 0.0001,
443
+ "loss": 0.0025,
444
+ "step": 620
445
+ },
446
+ {
447
+ "epoch": 2.52,
448
+ "grad_norm": 0.009542998857796192,
449
+ "learning_rate": 0.0001,
450
+ "loss": 0.0025,
451
+ "step": 630
452
+ },
453
+ {
454
+ "epoch": 2.56,
455
+ "grad_norm": 0.008792417123913765,
456
+ "learning_rate": 0.0001,
457
+ "loss": 0.0024,
458
+ "step": 640
459
+ },
460
+ {
461
+ "epoch": 2.6,
462
+ "grad_norm": 0.008578693494200706,
463
+ "learning_rate": 0.0001,
464
+ "loss": 0.0025,
465
+ "step": 650
466
+ },
467
+ {
468
+ "epoch": 2.64,
469
+ "grad_norm": 0.00964912585914135,
470
+ "learning_rate": 0.0001,
471
+ "loss": 0.0023,
472
+ "step": 660
473
+ },
474
+ {
475
+ "epoch": 2.68,
476
+ "grad_norm": 0.010405980981886387,
477
+ "learning_rate": 0.0001,
478
+ "loss": 0.0025,
479
+ "step": 670
480
+ },
481
+ {
482
+ "epoch": 2.7199999999999998,
483
+ "grad_norm": 0.010341090150177479,
484
+ "learning_rate": 0.0001,
485
+ "loss": 0.0025,
486
+ "step": 680
487
+ },
488
+ {
489
+ "epoch": 2.76,
490
+ "grad_norm": 0.008614353835582733,
491
+ "learning_rate": 0.0001,
492
+ "loss": 0.0025,
493
+ "step": 690
494
+ },
495
+ {
496
+ "epoch": 2.8,
497
+ "grad_norm": 0.007937280461192131,
498
+ "learning_rate": 0.0001,
499
+ "loss": 0.0023,
500
+ "step": 700
501
+ },
502
+ {
503
+ "epoch": 2.84,
504
+ "grad_norm": 0.01031468715518713,
505
+ "learning_rate": 0.0001,
506
+ "loss": 0.0023,
507
+ "step": 710
508
+ },
509
+ {
510
+ "epoch": 2.88,
511
+ "grad_norm": 0.00910355243831873,
512
+ "learning_rate": 0.0001,
513
+ "loss": 0.0022,
514
+ "step": 720
515
+ },
516
+ {
517
+ "epoch": 2.92,
518
+ "grad_norm": 0.00943271815776825,
519
+ "learning_rate": 0.0001,
520
+ "loss": 0.0023,
521
+ "step": 730
522
+ },
523
+ {
524
+ "epoch": 2.96,
525
+ "grad_norm": 0.009353592060506344,
526
+ "learning_rate": 0.0001,
527
+ "loss": 0.0022,
528
+ "step": 740
529
+ },
530
+ {
531
+ "epoch": 3.0,
532
+ "grad_norm": 0.010902035981416702,
533
+ "learning_rate": 0.0001,
534
+ "loss": 0.0023,
535
+ "step": 750
536
+ },
537
+ {
538
+ "epoch": 3.04,
539
+ "grad_norm": 0.009516863152384758,
540
+ "learning_rate": 0.0001,
541
+ "loss": 0.0021,
542
+ "step": 760
543
+ },
544
+ {
545
+ "epoch": 3.08,
546
+ "grad_norm": 0.011545202694833279,
547
+ "learning_rate": 0.0001,
548
+ "loss": 0.0021,
549
+ "step": 770
550
+ },
551
+ {
552
+ "epoch": 3.12,
553
+ "grad_norm": 0.007929637096822262,
554
+ "learning_rate": 0.0001,
555
+ "loss": 0.0021,
556
+ "step": 780
557
+ },
558
+ {
559
+ "epoch": 3.16,
560
+ "grad_norm": 0.008281239308416843,
561
+ "learning_rate": 0.0001,
562
+ "loss": 0.002,
563
+ "step": 790
564
+ },
565
+ {
566
+ "epoch": 3.2,
567
+ "grad_norm": 0.010873930528759956,
568
+ "learning_rate": 0.0001,
569
+ "loss": 0.0022,
570
+ "step": 800
571
+ },
572
+ {
573
+ "epoch": 3.24,
574
+ "grad_norm": 0.008918201550841331,
575
+ "learning_rate": 0.0001,
576
+ "loss": 0.0021,
577
+ "step": 810
578
+ },
579
+ {
580
+ "epoch": 3.2800000000000002,
581
+ "grad_norm": 0.010175652801990509,
582
+ "learning_rate": 0.0001,
583
+ "loss": 0.002,
584
+ "step": 820
585
+ },
586
+ {
587
+ "epoch": 3.32,
588
+ "grad_norm": 0.010831949301064014,
589
+ "learning_rate": 0.0001,
590
+ "loss": 0.002,
591
+ "step": 830
592
+ },
593
+ {
594
+ "epoch": 3.36,
595
+ "grad_norm": 0.008530331775546074,
596
+ "learning_rate": 0.0001,
597
+ "loss": 0.002,
598
+ "step": 840
599
+ },
600
+ {
601
+ "epoch": 3.4,
602
+ "grad_norm": 0.009083152748644352,
603
+ "learning_rate": 0.0001,
604
+ "loss": 0.002,
605
+ "step": 850
606
+ },
607
+ {
608
+ "epoch": 3.44,
609
+ "grad_norm": 0.009484045207500458,
610
+ "learning_rate": 0.0001,
611
+ "loss": 0.002,
612
+ "step": 860
613
+ },
614
+ {
615
+ "epoch": 3.48,
616
+ "grad_norm": 0.008554712869226933,
617
+ "learning_rate": 0.0001,
618
+ "loss": 0.0019,
619
+ "step": 870
620
+ },
621
+ {
622
+ "epoch": 3.52,
623
+ "grad_norm": 0.010989188216626644,
624
+ "learning_rate": 0.0001,
625
+ "loss": 0.0019,
626
+ "step": 880
627
+ },
628
+ {
629
+ "epoch": 3.56,
630
+ "grad_norm": 0.009386119432747364,
631
+ "learning_rate": 0.0001,
632
+ "loss": 0.0019,
633
+ "step": 890
634
+ },
635
+ {
636
+ "epoch": 3.6,
637
+ "grad_norm": 0.007983546704053879,
638
+ "learning_rate": 0.0001,
639
+ "loss": 0.002,
640
+ "step": 900
641
+ },
642
+ {
643
+ "epoch": 3.64,
644
+ "grad_norm": 0.008688435889780521,
645
+ "learning_rate": 0.0001,
646
+ "loss": 0.002,
647
+ "step": 910
648
+ },
649
+ {
650
+ "epoch": 3.68,
651
+ "grad_norm": 0.00950028095394373,
652
+ "learning_rate": 0.0001,
653
+ "loss": 0.0019,
654
+ "step": 920
655
+ },
656
+ {
657
+ "epoch": 3.7199999999999998,
658
+ "grad_norm": 0.010507030412554741,
659
+ "learning_rate": 0.0001,
660
+ "loss": 0.002,
661
+ "step": 930
662
+ },
663
+ {
664
+ "epoch": 3.76,
665
+ "grad_norm": 0.009756983257830143,
666
+ "learning_rate": 0.0001,
667
+ "loss": 0.0019,
668
+ "step": 940
669
+ },
670
+ {
671
+ "epoch": 3.8,
672
+ "grad_norm": 0.008237377740442753,
673
+ "learning_rate": 0.0001,
674
+ "loss": 0.0019,
675
+ "step": 950
676
+ },
677
+ {
678
+ "epoch": 3.84,
679
+ "grad_norm": 0.0092322314158082,
680
+ "learning_rate": 0.0001,
681
+ "loss": 0.002,
682
+ "step": 960
683
+ },
684
+ {
685
+ "epoch": 3.88,
686
+ "grad_norm": 0.006983477156609297,
687
+ "learning_rate": 0.0001,
688
+ "loss": 0.0019,
689
+ "step": 970
690
+ },
691
+ {
692
+ "epoch": 3.92,
693
+ "grad_norm": 0.009244760498404503,
694
+ "learning_rate": 0.0001,
695
+ "loss": 0.0019,
696
+ "step": 980
697
+ },
698
+ {
699
+ "epoch": 3.96,
700
+ "grad_norm": 0.009358828887343407,
701
+ "learning_rate": 0.0001,
702
+ "loss": 0.0018,
703
+ "step": 990
704
+ },
705
+ {
706
+ "epoch": 4.0,
707
+ "grad_norm": 0.009272817522287369,
708
+ "learning_rate": 0.0001,
709
+ "loss": 0.0018,
710
+ "step": 1000
711
+ },
712
+ {
713
+ "epoch": 4.0,
714
+ "step": 1000,
715
+ "total_flos": 1.9243641771917312e+18,
716
+ "train_loss": 0.050279883230105044,
717
+ "train_runtime": 7527.0065,
718
+ "train_samples_per_second": 21.257,
719
+ "train_steps_per_second": 0.133
720
  }
721
  ],
722
+ "logging_steps": 10,
723
+ "max_steps": 1000,
724
  "num_input_tokens_seen": 0,
725
+ "num_train_epochs": 4,
726
  "save_steps": 500,
727
  "stateful_callbacks": {
728
  "TrainerControl": {
 
736
  "attributes": {}
737
  }
738
  },
739
+ "total_flos": 1.9243641771917312e+18,
740
  "train_batch_size": 2,
741
  "trial_name": null,
742
  "trial_params": null
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0b4ef0ada0bb7702fd24371930da3d2298189a4d27c86b3c9aac6c85f106d575
3
- size 6289
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2d1e2f02fb5745ba0f4e76e2eff21d03ad5d85b29be9ff6ff5010da0e9f41530
3
+ size 9528
vocab.json ADDED
The diff for this file is too large to render. See raw diff