ksjpswaroop commited on
Commit
e127773
·
verified ·
1 Parent(s): ce20acd

Upload folder using huggingface_hub

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +5 -0
  2. checkpoint-1500/config.json +36 -0
  3. checkpoint-1500/generation_config.json +9 -0
  4. checkpoint-1500/model.safetensors +3 -0
  5. checkpoint-1500/optimizer.pt +3 -0
  6. checkpoint-1500/rng_state_0.pth +3 -0
  7. checkpoint-1500/rng_state_1.pth +3 -0
  8. checkpoint-1500/rng_state_2.pth +3 -0
  9. checkpoint-1500/rng_state_3.pth +3 -0
  10. checkpoint-1500/scheduler.pt +3 -0
  11. checkpoint-1500/tokenizer.json +3 -0
  12. checkpoint-1500/tokenizer_config.json +14 -0
  13. checkpoint-1500/trainer_state.json +244 -0
  14. checkpoint-1500/training_args.bin +3 -0
  15. checkpoint-2000/config.json +36 -0
  16. checkpoint-2000/generation_config.json +9 -0
  17. checkpoint-2000/model.safetensors +3 -0
  18. checkpoint-2000/optimizer.pt +3 -0
  19. checkpoint-2000/rng_state_0.pth +3 -0
  20. checkpoint-2000/rng_state_1.pth +3 -0
  21. checkpoint-2000/rng_state_2.pth +3 -0
  22. checkpoint-2000/rng_state_3.pth +3 -0
  23. checkpoint-2000/scheduler.pt +3 -0
  24. checkpoint-2000/tokenizer.json +3 -0
  25. checkpoint-2000/tokenizer_config.json +14 -0
  26. checkpoint-2000/trainer_state.json +314 -0
  27. checkpoint-2000/training_args.bin +3 -0
  28. checkpoint-2500/config.json +36 -0
  29. checkpoint-2500/generation_config.json +9 -0
  30. checkpoint-2500/model.safetensors +3 -0
  31. checkpoint-2500/optimizer.pt +3 -0
  32. checkpoint-2500/rng_state_0.pth +3 -0
  33. checkpoint-2500/rng_state_1.pth +3 -0
  34. checkpoint-2500/rng_state_2.pth +3 -0
  35. checkpoint-2500/rng_state_3.pth +3 -0
  36. checkpoint-2500/scheduler.pt +3 -0
  37. checkpoint-2500/tokenizer.json +3 -0
  38. checkpoint-2500/tokenizer_config.json +14 -0
  39. checkpoint-2500/trainer_state.json +384 -0
  40. checkpoint-2500/training_args.bin +3 -0
  41. checkpoint-3000/config.json +36 -0
  42. checkpoint-3000/generation_config.json +9 -0
  43. checkpoint-3000/model.safetensors +3 -0
  44. checkpoint-3000/optimizer.pt +3 -0
  45. checkpoint-3000/rng_state_0.pth +3 -0
  46. checkpoint-3000/rng_state_1.pth +3 -0
  47. checkpoint-3000/rng_state_2.pth +3 -0
  48. checkpoint-3000/rng_state_3.pth +3 -0
  49. checkpoint-3000/scheduler.pt +3 -0
  50. checkpoint-3000/tokenizer.json +3 -0
.gitattributes CHANGED
@@ -33,3 +33,8 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ checkpoint-1500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
37
+ checkpoint-2000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
38
+ checkpoint-2500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
39
+ checkpoint-3000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
40
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
checkpoint-1500/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LlamaForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 128000,
8
+ "dtype": "bfloat16",
9
+ "eos_token_id": 128001,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 3072,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 8192,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "num_attention_heads": 24,
19
+ "num_hidden_layers": 28,
20
+ "num_key_value_heads": 8,
21
+ "pad_token_id": null,
22
+ "pretraining_tp": 1,
23
+ "rms_norm_eps": 1e-05,
24
+ "rope_parameters": {
25
+ "factor": 32.0,
26
+ "high_freq_factor": 4.0,
27
+ "low_freq_factor": 1.0,
28
+ "original_max_position_embeddings": 8192,
29
+ "rope_theta": 500000.0,
30
+ "rope_type": "llama3"
31
+ },
32
+ "tie_word_embeddings": true,
33
+ "transformers_version": "5.2.0",
34
+ "use_cache": false,
35
+ "vocab_size": 128256
36
+ }
checkpoint-1500/generation_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 128000,
4
+ "do_sample": true,
5
+ "eos_token_id": 128001,
6
+ "temperature": 0.6,
7
+ "top_p": 0.9,
8
+ "transformers_version": "5.2.0"
9
+ }
checkpoint-1500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e03f7b409c3a292cb0f236a71344ba039503e62c2b9a429be411a734bb56be7f
3
+ size 6425529112
checkpoint-1500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fa02c88a1a885e03f7d46b34773567c9b6d16da58f6fe594bcdf0971a268273c
3
+ size 12851221159
checkpoint-1500/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b80829a78a755ba0cf5e985f9212f164229ce25f7c02fa877b4cdebf230922b5
3
+ size 15429
checkpoint-1500/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0c6f4074cb5241eff8959c1f45b18c23c893119f6c03fa2da27359dbe7dba66c
3
+ size 15429
checkpoint-1500/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f1e84026d823a1c04897f710cd119a65b654c08ad443e97c4486fc7a26f6e64b
3
+ size 15429
checkpoint-1500/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d9b60a41c989678f542cc75416e61e7a43f81d6221d259490134253d1f38a920
3
+ size 15429
checkpoint-1500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:64e8b2bdcac5ca2a3c5198f5b93e2fbd722a6494563b05b44a3f9f428774cad4
3
+ size 1465
checkpoint-1500/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b
3
+ size 17209920
checkpoint-1500/tokenizer_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "bos_token": "<|begin_of_text|>",
4
+ "clean_up_tokenization_spaces": true,
5
+ "eos_token": "<|end_of_text|>",
6
+ "is_local": true,
7
+ "model_input_names": [
8
+ "input_ids",
9
+ "attention_mask"
10
+ ],
11
+ "model_max_length": 131072,
12
+ "pad_token": "<|end_of_text|>",
13
+ "tokenizer_class": "TokenizersBackend"
14
+ }
checkpoint-1500/trainer_state.json ADDED
@@ -0,0 +1,244 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 4.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1500,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.13333333333333333,
14
+ "grad_norm": 1.4453125,
15
+ "learning_rate": 9.8e-06,
16
+ "loss": 1.4362579345703126,
17
+ "step": 50
18
+ },
19
+ {
20
+ "epoch": 0.26666666666666666,
21
+ "grad_norm": 1.109375,
22
+ "learning_rate": 1.98e-05,
23
+ "loss": 1.2353227996826173,
24
+ "step": 100
25
+ },
26
+ {
27
+ "epoch": 0.4,
28
+ "grad_norm": 1.359375,
29
+ "learning_rate": 2.98e-05,
30
+ "loss": 1.1962551879882812,
31
+ "step": 150
32
+ },
33
+ {
34
+ "epoch": 0.5333333333333333,
35
+ "grad_norm": 1.0,
36
+ "learning_rate": 2.997812450643705e-05,
37
+ "loss": 1.1289045715332031,
38
+ "step": 200
39
+ },
40
+ {
41
+ "epoch": 0.6666666666666666,
42
+ "grad_norm": 1.1640625,
43
+ "learning_rate": 2.991077005423577e-05,
44
+ "loss": 1.1314387512207031,
45
+ "step": 250
46
+ },
47
+ {
48
+ "epoch": 0.8,
49
+ "grad_norm": 1.140625,
50
+ "learning_rate": 2.979813208344657e-05,
51
+ "loss": 1.1489389038085938,
52
+ "step": 300
53
+ },
54
+ {
55
+ "epoch": 0.9333333333333333,
56
+ "grad_norm": 1.0703125,
57
+ "learning_rate": 2.964055267188472e-05,
58
+ "loss": 1.1160533142089843,
59
+ "step": 350
60
+ },
61
+ {
62
+ "epoch": 1.0666666666666667,
63
+ "grad_norm": 1.0625,
64
+ "learning_rate": 2.943851038304401e-05,
65
+ "loss": 1.040457763671875,
66
+ "step": 400
67
+ },
68
+ {
69
+ "epoch": 1.2,
70
+ "grad_norm": 1.140625,
71
+ "learning_rate": 2.9192618812715216e-05,
72
+ "loss": 0.924019775390625,
73
+ "step": 450
74
+ },
75
+ {
76
+ "epoch": 1.3333333333333333,
77
+ "grad_norm": 1.15625,
78
+ "learning_rate": 2.8903624725515786e-05,
79
+ "loss": 0.8882070922851563,
80
+ "step": 500
81
+ },
82
+ {
83
+ "epoch": 1.4666666666666668,
84
+ "grad_norm": 1.140625,
85
+ "learning_rate": 2.8572405786990293e-05,
86
+ "loss": 0.8971571350097656,
87
+ "step": 550
88
+ },
89
+ {
90
+ "epoch": 1.6,
91
+ "grad_norm": 1.1015625,
92
+ "learning_rate": 2.819996789816898e-05,
93
+ "loss": 0.9332583618164062,
94
+ "step": 600
95
+ },
96
+ {
97
+ "epoch": 1.7333333333333334,
98
+ "grad_norm": 1.203125,
99
+ "learning_rate": 2.7787442140679422e-05,
100
+ "loss": 0.8832476806640625,
101
+ "step": 650
102
+ },
103
+ {
104
+ "epoch": 1.8666666666666667,
105
+ "grad_norm": 1.2578125,
106
+ "learning_rate": 2.7336081341688787e-05,
107
+ "loss": 0.9302313232421875,
108
+ "step": 700
109
+ },
110
+ {
111
+ "epoch": 2.0,
112
+ "grad_norm": 1.15625,
113
+ "learning_rate": 2.684725626910889e-05,
114
+ "loss": 0.9123858642578125,
115
+ "step": 750
116
+ },
117
+ {
118
+ "epoch": 2.1333333333333333,
119
+ "grad_norm": 1.34375,
120
+ "learning_rate": 2.632245146861912e-05,
121
+ "loss": 0.6757525634765625,
122
+ "step": 800
123
+ },
124
+ {
125
+ "epoch": 2.2666666666666666,
126
+ "grad_norm": 1.4296875,
127
+ "learning_rate": 2.5763260755150013e-05,
128
+ "loss": 0.6670223999023438,
129
+ "step": 850
130
+ },
131
+ {
132
+ "epoch": 2.4,
133
+ "grad_norm": 1.34375,
134
+ "learning_rate": 2.5171382372519647e-05,
135
+ "loss": 0.6872428894042969,
136
+ "step": 900
137
+ },
138
+ {
139
+ "epoch": 2.533333333333333,
140
+ "grad_norm": 1.2109375,
141
+ "learning_rate": 2.4548613835923026e-05,
142
+ "loss": 0.6917278289794921,
143
+ "step": 950
144
+ },
145
+ {
146
+ "epoch": 2.6666666666666665,
147
+ "grad_norm": 1.25,
148
+ "learning_rate": 2.389684647293749e-05,
149
+ "loss": 0.6836544036865234,
150
+ "step": 1000
151
+ },
152
+ {
153
+ "epoch": 2.8,
154
+ "grad_norm": 1.5625,
155
+ "learning_rate": 2.3218059679623026e-05,
156
+ "loss": 0.6740491485595703,
157
+ "step": 1050
158
+ },
159
+ {
160
+ "epoch": 2.9333333333333336,
161
+ "grad_norm": 1.265625,
162
+ "learning_rate": 2.2514314909161486e-05,
163
+ "loss": 0.6897286987304687,
164
+ "step": 1100
165
+ },
166
+ {
167
+ "epoch": 3.066666666666667,
168
+ "grad_norm": 1.53125,
169
+ "learning_rate": 2.1787749411291056e-05,
170
+ "loss": 0.6019991302490234,
171
+ "step": 1150
172
+ },
173
+ {
174
+ "epoch": 3.2,
175
+ "grad_norm": 1.578125,
176
+ "learning_rate": 2.1040569741549002e-05,
177
+ "loss": 0.49853515625,
178
+ "step": 1200
179
+ },
180
+ {
181
+ "epoch": 3.3333333333333335,
182
+ "grad_norm": 1.625,
183
+ "learning_rate": 2.0275045060035044e-05,
184
+ "loss": 0.5089201354980468,
185
+ "step": 1250
186
+ },
187
+ {
188
+ "epoch": 3.466666666666667,
189
+ "grad_norm": 1.5234375,
190
+ "learning_rate": 1.9493500240046727e-05,
191
+ "loss": 0.5025421142578125,
192
+ "step": 1300
193
+ },
194
+ {
195
+ "epoch": 3.6,
196
+ "grad_norm": 1.3203125,
197
+ "learning_rate": 1.8698308807515613e-05,
198
+ "loss": 0.5206479263305664,
199
+ "step": 1350
200
+ },
201
+ {
202
+ "epoch": 3.7333333333333334,
203
+ "grad_norm": 1.4140625,
204
+ "learning_rate": 1.7891885732687026e-05,
205
+ "loss": 0.5091935729980469,
206
+ "step": 1400
207
+ },
208
+ {
209
+ "epoch": 3.8666666666666667,
210
+ "grad_norm": 1.3359375,
211
+ "learning_rate": 1.7076680095934813e-05,
212
+ "loss": 0.5084111785888672,
213
+ "step": 1450
214
+ },
215
+ {
216
+ "epoch": 4.0,
217
+ "grad_norm": 1.390625,
218
+ "learning_rate": 1.6255167649984738e-05,
219
+ "loss": 0.522473373413086,
220
+ "step": 1500
221
+ }
222
+ ],
223
+ "logging_steps": 50,
224
+ "max_steps": 3000,
225
+ "num_input_tokens_seen": 0,
226
+ "num_train_epochs": 8,
227
+ "save_steps": 500,
228
+ "stateful_callbacks": {
229
+ "TrainerControl": {
230
+ "args": {
231
+ "should_epoch_stop": false,
232
+ "should_evaluate": false,
233
+ "should_log": false,
234
+ "should_save": true,
235
+ "should_training_stop": false
236
+ },
237
+ "attributes": {}
238
+ }
239
+ },
240
+ "total_flos": 1.6625649181104538e+18,
241
+ "train_batch_size": 4,
242
+ "trial_name": null,
243
+ "trial_params": null
244
+ }
checkpoint-1500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9507d637355de62eba1ed595b47f14c9f1b33d51e48dea4351d68be1dc0fe373
3
+ size 5265
checkpoint-2000/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LlamaForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 128000,
8
+ "dtype": "bfloat16",
9
+ "eos_token_id": 128001,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 3072,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 8192,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "num_attention_heads": 24,
19
+ "num_hidden_layers": 28,
20
+ "num_key_value_heads": 8,
21
+ "pad_token_id": null,
22
+ "pretraining_tp": 1,
23
+ "rms_norm_eps": 1e-05,
24
+ "rope_parameters": {
25
+ "factor": 32.0,
26
+ "high_freq_factor": 4.0,
27
+ "low_freq_factor": 1.0,
28
+ "original_max_position_embeddings": 8192,
29
+ "rope_theta": 500000.0,
30
+ "rope_type": "llama3"
31
+ },
32
+ "tie_word_embeddings": true,
33
+ "transformers_version": "5.2.0",
34
+ "use_cache": false,
35
+ "vocab_size": 128256
36
+ }
checkpoint-2000/generation_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 128000,
4
+ "do_sample": true,
5
+ "eos_token_id": 128001,
6
+ "temperature": 0.6,
7
+ "top_p": 0.9,
8
+ "transformers_version": "5.2.0"
9
+ }
checkpoint-2000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:23fb39373725541de20531b434546ae44e6cc80545d1fc9d856f22204914e984
3
+ size 6425529112
checkpoint-2000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0960b3ed5c9d7a7e0e469d0963a7f780aedfa6c6fc8c8fc04fbe12bd829792b6
3
+ size 12851221159
checkpoint-2000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:485c3eb7b8ebe39984117ce19c256db4c11dba6dcd8d32908be6d94ca7a4d529
3
+ size 15429
checkpoint-2000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f7cd0007903b1a0344b4ba578f518e1366c3607febc6b70080704119da49e8fc
3
+ size 15429
checkpoint-2000/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a453120913ecd5657ebaf5fb69504cf1a158e4ff4015f66acf4be6e06bc270e0
3
+ size 15429
checkpoint-2000/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8c18edf48cc3adf4eb70b31f8e96972c6e25718fd8fcd0866b9ace124a9269d2
3
+ size 15429
checkpoint-2000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8f186aef94d351d54ba2d4a2a29efca5c57520f3b20fd57991689d6221f1607f
3
+ size 1465
checkpoint-2000/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b
3
+ size 17209920
checkpoint-2000/tokenizer_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "bos_token": "<|begin_of_text|>",
4
+ "clean_up_tokenization_spaces": true,
5
+ "eos_token": "<|end_of_text|>",
6
+ "is_local": true,
7
+ "model_input_names": [
8
+ "input_ids",
9
+ "attention_mask"
10
+ ],
11
+ "model_max_length": 131072,
12
+ "pad_token": "<|end_of_text|>",
13
+ "tokenizer_class": "TokenizersBackend"
14
+ }
checkpoint-2000/trainer_state.json ADDED
@@ -0,0 +1,314 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 5.333333333333333,
6
+ "eval_steps": 500,
7
+ "global_step": 2000,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.13333333333333333,
14
+ "grad_norm": 1.4453125,
15
+ "learning_rate": 9.8e-06,
16
+ "loss": 1.4362579345703126,
17
+ "step": 50
18
+ },
19
+ {
20
+ "epoch": 0.26666666666666666,
21
+ "grad_norm": 1.109375,
22
+ "learning_rate": 1.98e-05,
23
+ "loss": 1.2353227996826173,
24
+ "step": 100
25
+ },
26
+ {
27
+ "epoch": 0.4,
28
+ "grad_norm": 1.359375,
29
+ "learning_rate": 2.98e-05,
30
+ "loss": 1.1962551879882812,
31
+ "step": 150
32
+ },
33
+ {
34
+ "epoch": 0.5333333333333333,
35
+ "grad_norm": 1.0,
36
+ "learning_rate": 2.997812450643705e-05,
37
+ "loss": 1.1289045715332031,
38
+ "step": 200
39
+ },
40
+ {
41
+ "epoch": 0.6666666666666666,
42
+ "grad_norm": 1.1640625,
43
+ "learning_rate": 2.991077005423577e-05,
44
+ "loss": 1.1314387512207031,
45
+ "step": 250
46
+ },
47
+ {
48
+ "epoch": 0.8,
49
+ "grad_norm": 1.140625,
50
+ "learning_rate": 2.979813208344657e-05,
51
+ "loss": 1.1489389038085938,
52
+ "step": 300
53
+ },
54
+ {
55
+ "epoch": 0.9333333333333333,
56
+ "grad_norm": 1.0703125,
57
+ "learning_rate": 2.964055267188472e-05,
58
+ "loss": 1.1160533142089843,
59
+ "step": 350
60
+ },
61
+ {
62
+ "epoch": 1.0666666666666667,
63
+ "grad_norm": 1.0625,
64
+ "learning_rate": 2.943851038304401e-05,
65
+ "loss": 1.040457763671875,
66
+ "step": 400
67
+ },
68
+ {
69
+ "epoch": 1.2,
70
+ "grad_norm": 1.140625,
71
+ "learning_rate": 2.9192618812715216e-05,
72
+ "loss": 0.924019775390625,
73
+ "step": 450
74
+ },
75
+ {
76
+ "epoch": 1.3333333333333333,
77
+ "grad_norm": 1.15625,
78
+ "learning_rate": 2.8903624725515786e-05,
79
+ "loss": 0.8882070922851563,
80
+ "step": 500
81
+ },
82
+ {
83
+ "epoch": 1.4666666666666668,
84
+ "grad_norm": 1.140625,
85
+ "learning_rate": 2.8572405786990293e-05,
86
+ "loss": 0.8971571350097656,
87
+ "step": 550
88
+ },
89
+ {
90
+ "epoch": 1.6,
91
+ "grad_norm": 1.1015625,
92
+ "learning_rate": 2.819996789816898e-05,
93
+ "loss": 0.9332583618164062,
94
+ "step": 600
95
+ },
96
+ {
97
+ "epoch": 1.7333333333333334,
98
+ "grad_norm": 1.203125,
99
+ "learning_rate": 2.7787442140679422e-05,
100
+ "loss": 0.8832476806640625,
101
+ "step": 650
102
+ },
103
+ {
104
+ "epoch": 1.8666666666666667,
105
+ "grad_norm": 1.2578125,
106
+ "learning_rate": 2.7336081341688787e-05,
107
+ "loss": 0.9302313232421875,
108
+ "step": 700
109
+ },
110
+ {
111
+ "epoch": 2.0,
112
+ "grad_norm": 1.15625,
113
+ "learning_rate": 2.684725626910889e-05,
114
+ "loss": 0.9123858642578125,
115
+ "step": 750
116
+ },
117
+ {
118
+ "epoch": 2.1333333333333333,
119
+ "grad_norm": 1.34375,
120
+ "learning_rate": 2.632245146861912e-05,
121
+ "loss": 0.6757525634765625,
122
+ "step": 800
123
+ },
124
+ {
125
+ "epoch": 2.2666666666666666,
126
+ "grad_norm": 1.4296875,
127
+ "learning_rate": 2.5763260755150013e-05,
128
+ "loss": 0.6670223999023438,
129
+ "step": 850
130
+ },
131
+ {
132
+ "epoch": 2.4,
133
+ "grad_norm": 1.34375,
134
+ "learning_rate": 2.5171382372519647e-05,
135
+ "loss": 0.6872428894042969,
136
+ "step": 900
137
+ },
138
+ {
139
+ "epoch": 2.533333333333333,
140
+ "grad_norm": 1.2109375,
141
+ "learning_rate": 2.4548613835923026e-05,
142
+ "loss": 0.6917278289794921,
143
+ "step": 950
144
+ },
145
+ {
146
+ "epoch": 2.6666666666666665,
147
+ "grad_norm": 1.25,
148
+ "learning_rate": 2.389684647293749e-05,
149
+ "loss": 0.6836544036865234,
150
+ "step": 1000
151
+ },
152
+ {
153
+ "epoch": 2.8,
154
+ "grad_norm": 1.5625,
155
+ "learning_rate": 2.3218059679623026e-05,
156
+ "loss": 0.6740491485595703,
157
+ "step": 1050
158
+ },
159
+ {
160
+ "epoch": 2.9333333333333336,
161
+ "grad_norm": 1.265625,
162
+ "learning_rate": 2.2514314909161486e-05,
163
+ "loss": 0.6897286987304687,
164
+ "step": 1100
165
+ },
166
+ {
167
+ "epoch": 3.066666666666667,
168
+ "grad_norm": 1.53125,
169
+ "learning_rate": 2.1787749411291056e-05,
170
+ "loss": 0.6019991302490234,
171
+ "step": 1150
172
+ },
173
+ {
174
+ "epoch": 3.2,
175
+ "grad_norm": 1.578125,
176
+ "learning_rate": 2.1040569741549002e-05,
177
+ "loss": 0.49853515625,
178
+ "step": 1200
179
+ },
180
+ {
181
+ "epoch": 3.3333333333333335,
182
+ "grad_norm": 1.625,
183
+ "learning_rate": 2.0275045060035044e-05,
184
+ "loss": 0.5089201354980468,
185
+ "step": 1250
186
+ },
187
+ {
188
+ "epoch": 3.466666666666667,
189
+ "grad_norm": 1.5234375,
190
+ "learning_rate": 1.9493500240046727e-05,
191
+ "loss": 0.5025421142578125,
192
+ "step": 1300
193
+ },
194
+ {
195
+ "epoch": 3.6,
196
+ "grad_norm": 1.3203125,
197
+ "learning_rate": 1.8698308807515613e-05,
198
+ "loss": 0.5206479263305664,
199
+ "step": 1350
200
+ },
201
+ {
202
+ "epoch": 3.7333333333333334,
203
+ "grad_norm": 1.4140625,
204
+ "learning_rate": 1.7891885732687026e-05,
205
+ "loss": 0.5091935729980469,
206
+ "step": 1400
207
+ },
208
+ {
209
+ "epoch": 3.8666666666666667,
210
+ "grad_norm": 1.3359375,
211
+ "learning_rate": 1.7076680095934813e-05,
212
+ "loss": 0.5084111785888672,
213
+ "step": 1450
214
+ },
215
+ {
216
+ "epoch": 4.0,
217
+ "grad_norm": 1.390625,
218
+ "learning_rate": 1.6255167649984738e-05,
219
+ "loss": 0.522473373413086,
220
+ "step": 1500
221
+ },
222
+ {
223
+ "epoch": 4.133333333333334,
224
+ "grad_norm": 1.40625,
225
+ "learning_rate": 1.5429843301134782e-05,
226
+ "loss": 0.4226302337646484,
227
+ "step": 1550
228
+ },
229
+ {
230
+ "epoch": 4.266666666666667,
231
+ "grad_norm": 1.5078125,
232
+ "learning_rate": 1.4603213532306782e-05,
233
+ "loss": 0.40151473999023435,
234
+ "step": 1600
235
+ },
236
+ {
237
+ "epoch": 4.4,
238
+ "grad_norm": 1.1953125,
239
+ "learning_rate": 1.3777788790940136e-05,
240
+ "loss": 0.39469322204589846,
241
+ "step": 1650
242
+ },
243
+ {
244
+ "epoch": 4.533333333333333,
245
+ "grad_norm": 1.265625,
246
+ "learning_rate": 1.2956075864845435e-05,
247
+ "loss": 0.3755529022216797,
248
+ "step": 1700
249
+ },
250
+ {
251
+ "epoch": 4.666666666666667,
252
+ "grad_norm": 1.6953125,
253
+ "learning_rate": 1.21405702691723e-05,
254
+ "loss": 0.4053420257568359,
255
+ "step": 1750
256
+ },
257
+ {
258
+ "epoch": 4.8,
259
+ "grad_norm": 1.5859375,
260
+ "learning_rate": 1.1333748667611854e-05,
261
+ "loss": 0.3910691833496094,
262
+ "step": 1800
263
+ },
264
+ {
265
+ "epoch": 4.933333333333334,
266
+ "grad_norm": 1.515625,
267
+ "learning_rate": 1.0538061350850545e-05,
268
+ "loss": 0.41198493957519533,
269
+ "step": 1850
270
+ },
271
+ {
272
+ "epoch": 5.066666666666666,
273
+ "grad_norm": 1.3984375,
274
+ "learning_rate": 9.755924795117798e-06,
275
+ "loss": 0.3768700408935547,
276
+ "step": 1900
277
+ },
278
+ {
279
+ "epoch": 5.2,
280
+ "grad_norm": 1.5,
281
+ "learning_rate": 8.989714323427191e-06,
282
+ "loss": 0.3425804901123047,
283
+ "step": 1950
284
+ },
285
+ {
286
+ "epoch": 5.333333333333333,
287
+ "grad_norm": 1.4140625,
288
+ "learning_rate": 8.241756891798485e-06,
289
+ "loss": 0.3398160934448242,
290
+ "step": 2000
291
+ }
292
+ ],
293
+ "logging_steps": 50,
294
+ "max_steps": 3000,
295
+ "num_input_tokens_seen": 0,
296
+ "num_train_epochs": 8,
297
+ "save_steps": 500,
298
+ "stateful_callbacks": {
299
+ "TrainerControl": {
300
+ "args": {
301
+ "should_epoch_stop": false,
302
+ "should_evaluate": false,
303
+ "should_log": false,
304
+ "should_save": true,
305
+ "should_training_stop": false
306
+ },
307
+ "attributes": {}
308
+ }
309
+ },
310
+ "total_flos": 2.2167532241472717e+18,
311
+ "train_batch_size": 4,
312
+ "trial_name": null,
313
+ "trial_params": null
314
+ }
checkpoint-2000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9507d637355de62eba1ed595b47f14c9f1b33d51e48dea4351d68be1dc0fe373
3
+ size 5265
checkpoint-2500/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LlamaForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 128000,
8
+ "dtype": "bfloat16",
9
+ "eos_token_id": 128001,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 3072,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 8192,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "num_attention_heads": 24,
19
+ "num_hidden_layers": 28,
20
+ "num_key_value_heads": 8,
21
+ "pad_token_id": null,
22
+ "pretraining_tp": 1,
23
+ "rms_norm_eps": 1e-05,
24
+ "rope_parameters": {
25
+ "factor": 32.0,
26
+ "high_freq_factor": 4.0,
27
+ "low_freq_factor": 1.0,
28
+ "original_max_position_embeddings": 8192,
29
+ "rope_theta": 500000.0,
30
+ "rope_type": "llama3"
31
+ },
32
+ "tie_word_embeddings": true,
33
+ "transformers_version": "5.2.0",
34
+ "use_cache": false,
35
+ "vocab_size": 128256
36
+ }
checkpoint-2500/generation_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 128000,
4
+ "do_sample": true,
5
+ "eos_token_id": 128001,
6
+ "temperature": 0.6,
7
+ "top_p": 0.9,
8
+ "transformers_version": "5.2.0"
9
+ }
checkpoint-2500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ed27a08098f998beb90f01539b8e9287b3836104dc9aff236e338189998ba5f7
3
+ size 6425529112
checkpoint-2500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:10d0442d02feafd5414dd9b8da080f6d54a44fdac99ed2b78898bbb4e03346a6
3
+ size 12851221159
checkpoint-2500/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:63f8c8ac7017e076bdeab5686ee808dc34f2253a11672f9168162f1c139098a8
3
+ size 15429
checkpoint-2500/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:da9840d31d966ef4acd9553c28e732a2f97a00ae497036cfaf9f1c9aaf7bad46
3
+ size 15429
checkpoint-2500/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3c0d619a42d9d6b525fc902030d45e64a2643605df5c896eef51533128bd7def
3
+ size 15429
checkpoint-2500/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5dcc41326d60bf7e5dae8feb8751b9fe5898a4cea6e4b58c37c2d6f2164d12ea
3
+ size 15429
checkpoint-2500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f29dc8ae3cab366b02c0b5bf27f08a3f22095fddf6c5ecc87bdc43329fd253cb
3
+ size 1465
checkpoint-2500/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b
3
+ size 17209920
checkpoint-2500/tokenizer_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "bos_token": "<|begin_of_text|>",
4
+ "clean_up_tokenization_spaces": true,
5
+ "eos_token": "<|end_of_text|>",
6
+ "is_local": true,
7
+ "model_input_names": [
8
+ "input_ids",
9
+ "attention_mask"
10
+ ],
11
+ "model_max_length": 131072,
12
+ "pad_token": "<|end_of_text|>",
13
+ "tokenizer_class": "TokenizersBackend"
14
+ }
checkpoint-2500/trainer_state.json ADDED
@@ -0,0 +1,384 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 6.666666666666667,
6
+ "eval_steps": 500,
7
+ "global_step": 2500,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.13333333333333333,
14
+ "grad_norm": 1.4453125,
15
+ "learning_rate": 9.8e-06,
16
+ "loss": 1.4362579345703126,
17
+ "step": 50
18
+ },
19
+ {
20
+ "epoch": 0.26666666666666666,
21
+ "grad_norm": 1.109375,
22
+ "learning_rate": 1.98e-05,
23
+ "loss": 1.2353227996826173,
24
+ "step": 100
25
+ },
26
+ {
27
+ "epoch": 0.4,
28
+ "grad_norm": 1.359375,
29
+ "learning_rate": 2.98e-05,
30
+ "loss": 1.1962551879882812,
31
+ "step": 150
32
+ },
33
+ {
34
+ "epoch": 0.5333333333333333,
35
+ "grad_norm": 1.0,
36
+ "learning_rate": 2.997812450643705e-05,
37
+ "loss": 1.1289045715332031,
38
+ "step": 200
39
+ },
40
+ {
41
+ "epoch": 0.6666666666666666,
42
+ "grad_norm": 1.1640625,
43
+ "learning_rate": 2.991077005423577e-05,
44
+ "loss": 1.1314387512207031,
45
+ "step": 250
46
+ },
47
+ {
48
+ "epoch": 0.8,
49
+ "grad_norm": 1.140625,
50
+ "learning_rate": 2.979813208344657e-05,
51
+ "loss": 1.1489389038085938,
52
+ "step": 300
53
+ },
54
+ {
55
+ "epoch": 0.9333333333333333,
56
+ "grad_norm": 1.0703125,
57
+ "learning_rate": 2.964055267188472e-05,
58
+ "loss": 1.1160533142089843,
59
+ "step": 350
60
+ },
61
+ {
62
+ "epoch": 1.0666666666666667,
63
+ "grad_norm": 1.0625,
64
+ "learning_rate": 2.943851038304401e-05,
65
+ "loss": 1.040457763671875,
66
+ "step": 400
67
+ },
68
+ {
69
+ "epoch": 1.2,
70
+ "grad_norm": 1.140625,
71
+ "learning_rate": 2.9192618812715216e-05,
72
+ "loss": 0.924019775390625,
73
+ "step": 450
74
+ },
75
+ {
76
+ "epoch": 1.3333333333333333,
77
+ "grad_norm": 1.15625,
78
+ "learning_rate": 2.8903624725515786e-05,
79
+ "loss": 0.8882070922851563,
80
+ "step": 500
81
+ },
82
+ {
83
+ "epoch": 1.4666666666666668,
84
+ "grad_norm": 1.140625,
85
+ "learning_rate": 2.8572405786990293e-05,
86
+ "loss": 0.8971571350097656,
87
+ "step": 550
88
+ },
89
+ {
90
+ "epoch": 1.6,
91
+ "grad_norm": 1.1015625,
92
+ "learning_rate": 2.819996789816898e-05,
93
+ "loss": 0.9332583618164062,
94
+ "step": 600
95
+ },
96
+ {
97
+ "epoch": 1.7333333333333334,
98
+ "grad_norm": 1.203125,
99
+ "learning_rate": 2.7787442140679422e-05,
100
+ "loss": 0.8832476806640625,
101
+ "step": 650
102
+ },
103
+ {
104
+ "epoch": 1.8666666666666667,
105
+ "grad_norm": 1.2578125,
106
+ "learning_rate": 2.7336081341688787e-05,
107
+ "loss": 0.9302313232421875,
108
+ "step": 700
109
+ },
110
+ {
111
+ "epoch": 2.0,
112
+ "grad_norm": 1.15625,
113
+ "learning_rate": 2.684725626910889e-05,
114
+ "loss": 0.9123858642578125,
115
+ "step": 750
116
+ },
117
+ {
118
+ "epoch": 2.1333333333333333,
119
+ "grad_norm": 1.34375,
120
+ "learning_rate": 2.632245146861912e-05,
121
+ "loss": 0.6757525634765625,
122
+ "step": 800
123
+ },
124
+ {
125
+ "epoch": 2.2666666666666666,
126
+ "grad_norm": 1.4296875,
127
+ "learning_rate": 2.5763260755150013e-05,
128
+ "loss": 0.6670223999023438,
129
+ "step": 850
130
+ },
131
+ {
132
+ "epoch": 2.4,
133
+ "grad_norm": 1.34375,
134
+ "learning_rate": 2.5171382372519647e-05,
135
+ "loss": 0.6872428894042969,
136
+ "step": 900
137
+ },
138
+ {
139
+ "epoch": 2.533333333333333,
140
+ "grad_norm": 1.2109375,
141
+ "learning_rate": 2.4548613835923026e-05,
142
+ "loss": 0.6917278289794921,
143
+ "step": 950
144
+ },
145
+ {
146
+ "epoch": 2.6666666666666665,
147
+ "grad_norm": 1.25,
148
+ "learning_rate": 2.389684647293749e-05,
149
+ "loss": 0.6836544036865234,
150
+ "step": 1000
151
+ },
152
+ {
153
+ "epoch": 2.8,
154
+ "grad_norm": 1.5625,
155
+ "learning_rate": 2.3218059679623026e-05,
156
+ "loss": 0.6740491485595703,
157
+ "step": 1050
158
+ },
159
+ {
160
+ "epoch": 2.9333333333333336,
161
+ "grad_norm": 1.265625,
162
+ "learning_rate": 2.2514314909161486e-05,
163
+ "loss": 0.6897286987304687,
164
+ "step": 1100
165
+ },
166
+ {
167
+ "epoch": 3.066666666666667,
168
+ "grad_norm": 1.53125,
169
+ "learning_rate": 2.1787749411291056e-05,
170
+ "loss": 0.6019991302490234,
171
+ "step": 1150
172
+ },
173
+ {
174
+ "epoch": 3.2,
175
+ "grad_norm": 1.578125,
176
+ "learning_rate": 2.1040569741549002e-05,
177
+ "loss": 0.49853515625,
178
+ "step": 1200
179
+ },
180
+ {
181
+ "epoch": 3.3333333333333335,
182
+ "grad_norm": 1.625,
183
+ "learning_rate": 2.0275045060035044e-05,
184
+ "loss": 0.5089201354980468,
185
+ "step": 1250
186
+ },
187
+ {
188
+ "epoch": 3.466666666666667,
189
+ "grad_norm": 1.5234375,
190
+ "learning_rate": 1.9493500240046727e-05,
191
+ "loss": 0.5025421142578125,
192
+ "step": 1300
193
+ },
194
+ {
195
+ "epoch": 3.6,
196
+ "grad_norm": 1.3203125,
197
+ "learning_rate": 1.8698308807515613e-05,
198
+ "loss": 0.5206479263305664,
199
+ "step": 1350
200
+ },
201
+ {
202
+ "epoch": 3.7333333333333334,
203
+ "grad_norm": 1.4140625,
204
+ "learning_rate": 1.7891885732687026e-05,
205
+ "loss": 0.5091935729980469,
206
+ "step": 1400
207
+ },
208
+ {
209
+ "epoch": 3.8666666666666667,
210
+ "grad_norm": 1.3359375,
211
+ "learning_rate": 1.7076680095934813e-05,
212
+ "loss": 0.5084111785888672,
213
+ "step": 1450
214
+ },
215
+ {
216
+ "epoch": 4.0,
217
+ "grad_norm": 1.390625,
218
+ "learning_rate": 1.6255167649984738e-05,
219
+ "loss": 0.522473373413086,
220
+ "step": 1500
221
+ },
222
+ {
223
+ "epoch": 4.133333333333334,
224
+ "grad_norm": 1.40625,
225
+ "learning_rate": 1.5429843301134782e-05,
226
+ "loss": 0.4226302337646484,
227
+ "step": 1550
228
+ },
229
+ {
230
+ "epoch": 4.266666666666667,
231
+ "grad_norm": 1.5078125,
232
+ "learning_rate": 1.4603213532306782e-05,
233
+ "loss": 0.40151473999023435,
234
+ "step": 1600
235
+ },
236
+ {
237
+ "epoch": 4.4,
238
+ "grad_norm": 1.1953125,
239
+ "learning_rate": 1.3777788790940136e-05,
240
+ "loss": 0.39469322204589846,
241
+ "step": 1650
242
+ },
243
+ {
244
+ "epoch": 4.533333333333333,
245
+ "grad_norm": 1.265625,
246
+ "learning_rate": 1.2956075864845435e-05,
247
+ "loss": 0.3755529022216797,
248
+ "step": 1700
249
+ },
250
+ {
251
+ "epoch": 4.666666666666667,
252
+ "grad_norm": 1.6953125,
253
+ "learning_rate": 1.21405702691723e-05,
254
+ "loss": 0.4053420257568359,
255
+ "step": 1750
256
+ },
257
+ {
258
+ "epoch": 4.8,
259
+ "grad_norm": 1.5859375,
260
+ "learning_rate": 1.1333748667611854e-05,
261
+ "loss": 0.3910691833496094,
262
+ "step": 1800
263
+ },
264
+ {
265
+ "epoch": 4.933333333333334,
266
+ "grad_norm": 1.515625,
267
+ "learning_rate": 1.0538061350850545e-05,
268
+ "loss": 0.41198493957519533,
269
+ "step": 1850
270
+ },
271
+ {
272
+ "epoch": 5.066666666666666,
273
+ "grad_norm": 1.3984375,
274
+ "learning_rate": 9.755924795117798e-06,
275
+ "loss": 0.3768700408935547,
276
+ "step": 1900
277
+ },
278
+ {
279
+ "epoch": 5.2,
280
+ "grad_norm": 1.5,
281
+ "learning_rate": 8.989714323427191e-06,
282
+ "loss": 0.3425804901123047,
283
+ "step": 1950
284
+ },
285
+ {
286
+ "epoch": 5.333333333333333,
287
+ "grad_norm": 1.4140625,
288
+ "learning_rate": 8.241756891798485e-06,
289
+ "loss": 0.3398160934448242,
290
+ "step": 2000
291
+ },
292
+ {
293
+ "epoch": 5.466666666666667,
294
+ "grad_norm": 1.953125,
295
+ "learning_rate": 7.5143240223686934e-06,
296
+ "loss": 0.33743438720703123,
297
+ "step": 2050
298
+ },
299
+ {
300
+ "epoch": 5.6,
301
+ "grad_norm": 1.2421875,
302
+ "learning_rate": 6.809624904854006e-06,
303
+ "loss": 0.3318545532226562,
304
+ "step": 2100
305
+ },
306
+ {
307
+ "epoch": 5.733333333333333,
308
+ "grad_norm": 1.6171875,
309
+ "learning_rate": 6.12979968731323e-06,
310
+ "loss": 0.33765140533447263,
311
+ "step": 2150
312
+ },
313
+ {
314
+ "epoch": 5.866666666666667,
315
+ "grad_norm": 1.5703125,
316
+ "learning_rate": 5.4769129765884716e-06,
317
+ "loss": 0.34544857025146486,
318
+ "step": 2200
319
+ },
320
+ {
321
+ "epoch": 6.0,
322
+ "grad_norm": 1.4453125,
323
+ "learning_rate": 4.852947568162011e-06,
324
+ "loss": 0.34912044525146485,
325
+ "step": 2250
326
+ },
327
+ {
328
+ "epoch": 6.133333333333334,
329
+ "grad_norm": 1.7109375,
330
+ "learning_rate": 4.2597984244716e-06,
331
+ "loss": 0.32130874633789064,
332
+ "step": 2300
333
+ },
334
+ {
335
+ "epoch": 6.266666666666667,
336
+ "grad_norm": 1.4296875,
337
+ "learning_rate": 3.699266919971811e-06,
338
+ "loss": 0.32198089599609375,
339
+ "step": 2350
340
+ },
341
+ {
342
+ "epoch": 6.4,
343
+ "grad_norm": 1.3125,
344
+ "learning_rate": 3.1730553704190283e-06,
345
+ "loss": 0.3156669425964356,
346
+ "step": 2400
347
+ },
348
+ {
349
+ "epoch": 6.533333333333333,
350
+ "grad_norm": 1.5078125,
351
+ "learning_rate": 2.6827618629944394e-06,
352
+ "loss": 0.33737548828125,
353
+ "step": 2450
354
+ },
355
+ {
356
+ "epoch": 6.666666666666667,
357
+ "grad_norm": 1.4609375,
358
+ "learning_rate": 2.229875402965799e-06,
359
+ "loss": 0.31845829010009763,
360
+ "step": 2500
361
+ }
362
+ ],
363
+ "logging_steps": 50,
364
+ "max_steps": 3000,
365
+ "num_input_tokens_seen": 0,
366
+ "num_train_epochs": 8,
367
+ "save_steps": 500,
368
+ "stateful_callbacks": {
369
+ "TrainerControl": {
370
+ "args": {
371
+ "should_epoch_stop": false,
372
+ "should_evaluate": false,
373
+ "should_log": false,
374
+ "should_save": true,
375
+ "should_training_stop": false
376
+ },
377
+ "attributes": {}
378
+ }
379
+ },
380
+ "total_flos": 2.7709415301840896e+18,
381
+ "train_batch_size": 4,
382
+ "trial_name": null,
383
+ "trial_params": null
384
+ }
checkpoint-2500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9507d637355de62eba1ed595b47f14c9f1b33d51e48dea4351d68be1dc0fe373
3
+ size 5265
checkpoint-3000/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LlamaForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 128000,
8
+ "dtype": "bfloat16",
9
+ "eos_token_id": 128001,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 3072,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 8192,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "num_attention_heads": 24,
19
+ "num_hidden_layers": 28,
20
+ "num_key_value_heads": 8,
21
+ "pad_token_id": null,
22
+ "pretraining_tp": 1,
23
+ "rms_norm_eps": 1e-05,
24
+ "rope_parameters": {
25
+ "factor": 32.0,
26
+ "high_freq_factor": 4.0,
27
+ "low_freq_factor": 1.0,
28
+ "original_max_position_embeddings": 8192,
29
+ "rope_theta": 500000.0,
30
+ "rope_type": "llama3"
31
+ },
32
+ "tie_word_embeddings": true,
33
+ "transformers_version": "5.2.0",
34
+ "use_cache": false,
35
+ "vocab_size": 128256
36
+ }
checkpoint-3000/generation_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 128000,
4
+ "do_sample": true,
5
+ "eos_token_id": 128001,
6
+ "temperature": 0.6,
7
+ "top_p": 0.9,
8
+ "transformers_version": "5.2.0"
9
+ }
checkpoint-3000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8b2150100a3b809762228e21f49ec6830e931cc5df587fb905d72a39af8b6388
3
+ size 6425529112
checkpoint-3000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ae6fc851701134a88146b2d47beaa55d122cb526daaeeb0ddbf387a4181a58b2
3
+ size 12851221159
checkpoint-3000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ccdd04b664fbfa5d05e90c688a91f18c4d07f58893b9b3a4cfad78d78efbc6ac
3
+ size 15429
checkpoint-3000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5a41e144bee8116d5a47773b5f326ca7bb65df8c457f0c58a2ed8003d7ffcc73
3
+ size 15429
checkpoint-3000/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fa38d5d663317603edb36bb18e908202a95a6e6250c06e2bea719fdba77f1277
3
+ size 15429
checkpoint-3000/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dbe1f72976acfa5e0e28d92c61725f779f476ad95d06ce9b63a6d9978dd7b9ba
3
+ size 15429
checkpoint-3000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:36922d6c505490e075f78f99bc9d4cf97ba7133355550051865a88ede5aaf5ff
3
+ size 1465
checkpoint-3000/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b
3
+ size 17209920