Technotech commited on
Commit
b5bfb33
1 Parent(s): 7fae12c

Initial upload

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. README.md +18 -1
  2. adapter_config.json +21 -0
  3. adapter_model.bin +3 -0
  4. checkpoint-12360/README.md +20 -0
  5. checkpoint-12360/adapter_config.json +21 -0
  6. checkpoint-12360/adapter_model.bin +3 -0
  7. checkpoint-12360/optimizer.pt +3 -0
  8. checkpoint-12360/rng_state.pth +3 -0
  9. checkpoint-12360/scheduler.pt +3 -0
  10. checkpoint-12360/trainer_state.json +754 -0
  11. checkpoint-12360/training_args.bin +3 -0
  12. checkpoint-12390/README.md +20 -0
  13. checkpoint-12390/adapter_config.json +21 -0
  14. checkpoint-12390/adapter_model.bin +3 -0
  15. checkpoint-12390/optimizer.pt +3 -0
  16. checkpoint-12390/rng_state.pth +3 -0
  17. checkpoint-12390/scheduler.pt +3 -0
  18. checkpoint-12390/trainer_state.json +754 -0
  19. checkpoint-12390/training_args.bin +3 -0
  20. checkpoint-12420/README.md +20 -0
  21. checkpoint-12420/adapter_config.json +21 -0
  22. checkpoint-12420/adapter_model.bin +3 -0
  23. checkpoint-12420/optimizer.pt +3 -0
  24. checkpoint-12420/rng_state.pth +3 -0
  25. checkpoint-12420/scheduler.pt +3 -0
  26. checkpoint-12420/trainer_state.json +760 -0
  27. checkpoint-12420/training_args.bin +3 -0
  28. checkpoint-12450/README.md +20 -0
  29. checkpoint-12450/adapter_config.json +21 -0
  30. checkpoint-12450/adapter_model.bin +3 -0
  31. checkpoint-12450/optimizer.pt +3 -0
  32. checkpoint-12450/rng_state.pth +3 -0
  33. checkpoint-12450/scheduler.pt +3 -0
  34. checkpoint-12450/trainer_state.json +760 -0
  35. checkpoint-12450/training_args.bin +3 -0
  36. checkpoint-12480/README.md +20 -0
  37. checkpoint-12480/adapter_config.json +21 -0
  38. checkpoint-12480/adapter_model.bin +3 -0
  39. checkpoint-12480/optimizer.pt +3 -0
  40. checkpoint-12480/rng_state.pth +3 -0
  41. checkpoint-12480/scheduler.pt +3 -0
  42. checkpoint-12480/trainer_state.json +760 -0
  43. checkpoint-12480/training_args.bin +3 -0
  44. checkpoint-12510/README.md +20 -0
  45. checkpoint-12510/adapter_config.json +21 -0
  46. checkpoint-12510/adapter_model.bin +3 -0
  47. checkpoint-12510/optimizer.pt +3 -0
  48. checkpoint-12510/rng_state.pth +3 -0
  49. checkpoint-12510/scheduler.pt +3 -0
  50. checkpoint-12510/trainer_state.json +766 -0
README.md CHANGED
@@ -1,3 +1,20 @@
1
  ---
2
- license: apache-2.0
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ library_name: peft
3
  ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: fp4
15
+ - bnb_4bit_use_double_quant: False
16
+ - bnb_4bit_compute_dtype: float32
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.5.0.dev0
adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "models/TinyStories-33M",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 8,
11
+ "lora_dropout": 0.0,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 16,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8c97a3bd6b22c34728f2f3fab2f809d43641be7294bd9b98909fc4f7fc889a64
3
+ size 792353
checkpoint-12360/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: fp4
15
+ - bnb_4bit_use_double_quant: False
16
+ - bnb_4bit_compute_dtype: float32
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.5.0.dev0
checkpoint-12360/adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "models/TinyStories-33M",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 8,
11
+ "lora_dropout": 0.0,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 16,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
checkpoint-12360/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f3bd39cae0272bd5fae91b640061201e810518807bc8c427109022a1099d8aac
3
+ size 792353
checkpoint-12360/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0ed27b4cc5ad7e5f3014aa32fc8c3cf435e1a448088270805ce3e38e082a5cd9
3
+ size 1586873
checkpoint-12360/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8bb3860d98363e44bc4704fb9458deec927fd96997d1d09c80639070db959d22
3
+ size 14575
checkpoint-12360/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:845c1f57b7b77abd8b8e3a8652626f29e87a4177c0fdccb7d2b228b260242406
3
+ size 627
checkpoint-12360/trainer_state.json ADDED
@@ -0,0 +1,754 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 9.65625,
5
+ "global_step": 12360,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.08,
12
+ "learning_rate": 0.0001,
13
+ "loss": 10.386,
14
+ "step": 100
15
+ },
16
+ {
17
+ "epoch": 0.16,
18
+ "learning_rate": 0.0001,
19
+ "loss": 7.4061,
20
+ "step": 200
21
+ },
22
+ {
23
+ "epoch": 0.23,
24
+ "learning_rate": 0.0001,
25
+ "loss": 6.4112,
26
+ "step": 300
27
+ },
28
+ {
29
+ "epoch": 0.31,
30
+ "learning_rate": 0.0001,
31
+ "loss": 6.091,
32
+ "step": 400
33
+ },
34
+ {
35
+ "epoch": 0.39,
36
+ "learning_rate": 0.0001,
37
+ "loss": 5.8571,
38
+ "step": 500
39
+ },
40
+ {
41
+ "epoch": 0.47,
42
+ "learning_rate": 0.0001,
43
+ "loss": 5.7144,
44
+ "step": 600
45
+ },
46
+ {
47
+ "epoch": 0.55,
48
+ "learning_rate": 0.0001,
49
+ "loss": 5.6228,
50
+ "step": 700
51
+ },
52
+ {
53
+ "epoch": 0.62,
54
+ "learning_rate": 0.0001,
55
+ "loss": 5.546,
56
+ "step": 800
57
+ },
58
+ {
59
+ "epoch": 0.7,
60
+ "learning_rate": 0.0001,
61
+ "loss": 5.4904,
62
+ "step": 900
63
+ },
64
+ {
65
+ "epoch": 0.78,
66
+ "learning_rate": 0.0001,
67
+ "loss": 5.4413,
68
+ "step": 1000
69
+ },
70
+ {
71
+ "epoch": 0.86,
72
+ "learning_rate": 0.0001,
73
+ "loss": 5.3603,
74
+ "step": 1100
75
+ },
76
+ {
77
+ "epoch": 0.94,
78
+ "learning_rate": 0.0001,
79
+ "loss": 5.3346,
80
+ "step": 1200
81
+ },
82
+ {
83
+ "epoch": 1.02,
84
+ "learning_rate": 0.0001,
85
+ "loss": 5.2744,
86
+ "step": 1300
87
+ },
88
+ {
89
+ "epoch": 1.09,
90
+ "learning_rate": 0.0001,
91
+ "loss": 5.231,
92
+ "step": 1400
93
+ },
94
+ {
95
+ "epoch": 1.17,
96
+ "learning_rate": 0.0001,
97
+ "loss": 5.1782,
98
+ "step": 1500
99
+ },
100
+ {
101
+ "epoch": 1.25,
102
+ "learning_rate": 0.0001,
103
+ "loss": 5.1613,
104
+ "step": 1600
105
+ },
106
+ {
107
+ "epoch": 1.33,
108
+ "learning_rate": 0.0001,
109
+ "loss": 5.1091,
110
+ "step": 1700
111
+ },
112
+ {
113
+ "epoch": 1.41,
114
+ "learning_rate": 0.0001,
115
+ "loss": 5.0939,
116
+ "step": 1800
117
+ },
118
+ {
119
+ "epoch": 1.48,
120
+ "learning_rate": 0.0001,
121
+ "loss": 5.0667,
122
+ "step": 1900
123
+ },
124
+ {
125
+ "epoch": 1.56,
126
+ "learning_rate": 0.0001,
127
+ "loss": 5.0054,
128
+ "step": 2000
129
+ },
130
+ {
131
+ "epoch": 1.64,
132
+ "learning_rate": 0.0001,
133
+ "loss": 4.96,
134
+ "step": 2100
135
+ },
136
+ {
137
+ "epoch": 1.72,
138
+ "learning_rate": 0.0001,
139
+ "loss": 4.9301,
140
+ "step": 2200
141
+ },
142
+ {
143
+ "epoch": 1.8,
144
+ "learning_rate": 0.0001,
145
+ "loss": 4.9302,
146
+ "step": 2300
147
+ },
148
+ {
149
+ "epoch": 1.88,
150
+ "learning_rate": 0.0001,
151
+ "loss": 4.9129,
152
+ "step": 2400
153
+ },
154
+ {
155
+ "epoch": 1.95,
156
+ "learning_rate": 0.0001,
157
+ "loss": 4.8627,
158
+ "step": 2500
159
+ },
160
+ {
161
+ "epoch": 2.03,
162
+ "learning_rate": 0.0001,
163
+ "loss": 4.8336,
164
+ "step": 2600
165
+ },
166
+ {
167
+ "epoch": 2.11,
168
+ "learning_rate": 0.0001,
169
+ "loss": 4.8018,
170
+ "step": 2700
171
+ },
172
+ {
173
+ "epoch": 2.19,
174
+ "learning_rate": 0.0001,
175
+ "loss": 4.7479,
176
+ "step": 2800
177
+ },
178
+ {
179
+ "epoch": 2.27,
180
+ "learning_rate": 0.0001,
181
+ "loss": 4.7572,
182
+ "step": 2900
183
+ },
184
+ {
185
+ "epoch": 2.34,
186
+ "learning_rate": 0.0001,
187
+ "loss": 4.759,
188
+ "step": 3000
189
+ },
190
+ {
191
+ "epoch": 2.42,
192
+ "learning_rate": 0.0001,
193
+ "loss": 4.7441,
194
+ "step": 3100
195
+ },
196
+ {
197
+ "epoch": 2.5,
198
+ "learning_rate": 0.0001,
199
+ "loss": 4.7195,
200
+ "step": 3200
201
+ },
202
+ {
203
+ "epoch": 2.58,
204
+ "learning_rate": 0.0001,
205
+ "loss": 4.6821,
206
+ "step": 3300
207
+ },
208
+ {
209
+ "epoch": 2.66,
210
+ "learning_rate": 0.0001,
211
+ "loss": 4.6587,
212
+ "step": 3400
213
+ },
214
+ {
215
+ "epoch": 2.73,
216
+ "learning_rate": 0.0001,
217
+ "loss": 4.6228,
218
+ "step": 3500
219
+ },
220
+ {
221
+ "epoch": 2.81,
222
+ "learning_rate": 0.0001,
223
+ "loss": 4.6248,
224
+ "step": 3600
225
+ },
226
+ {
227
+ "epoch": 2.89,
228
+ "learning_rate": 0.0001,
229
+ "loss": 4.5801,
230
+ "step": 3700
231
+ },
232
+ {
233
+ "epoch": 2.97,
234
+ "learning_rate": 0.0001,
235
+ "loss": 4.5934,
236
+ "step": 3800
237
+ },
238
+ {
239
+ "epoch": 3.05,
240
+ "learning_rate": 0.0001,
241
+ "loss": 4.5781,
242
+ "step": 3900
243
+ },
244
+ {
245
+ "epoch": 3.12,
246
+ "learning_rate": 0.0001,
247
+ "loss": 4.5763,
248
+ "step": 4000
249
+ },
250
+ {
251
+ "epoch": 3.2,
252
+ "learning_rate": 0.0001,
253
+ "loss": 4.5598,
254
+ "step": 4100
255
+ },
256
+ {
257
+ "epoch": 3.28,
258
+ "learning_rate": 0.0001,
259
+ "loss": 4.4796,
260
+ "step": 4200
261
+ },
262
+ {
263
+ "epoch": 3.36,
264
+ "learning_rate": 0.0001,
265
+ "loss": 4.5204,
266
+ "step": 4300
267
+ },
268
+ {
269
+ "epoch": 3.44,
270
+ "learning_rate": 0.0001,
271
+ "loss": 4.4858,
272
+ "step": 4400
273
+ },
274
+ {
275
+ "epoch": 3.52,
276
+ "learning_rate": 0.0001,
277
+ "loss": 4.4882,
278
+ "step": 4500
279
+ },
280
+ {
281
+ "epoch": 3.59,
282
+ "learning_rate": 0.0001,
283
+ "loss": 4.4852,
284
+ "step": 4600
285
+ },
286
+ {
287
+ "epoch": 3.67,
288
+ "learning_rate": 0.0001,
289
+ "loss": 4.4302,
290
+ "step": 4700
291
+ },
292
+ {
293
+ "epoch": 3.75,
294
+ "learning_rate": 0.0001,
295
+ "loss": 4.4267,
296
+ "step": 4800
297
+ },
298
+ {
299
+ "epoch": 3.83,
300
+ "learning_rate": 0.0001,
301
+ "loss": 4.4298,
302
+ "step": 4900
303
+ },
304
+ {
305
+ "epoch": 3.91,
306
+ "learning_rate": 0.0001,
307
+ "loss": 4.3892,
308
+ "step": 5000
309
+ },
310
+ {
311
+ "epoch": 3.98,
312
+ "learning_rate": 0.0001,
313
+ "loss": 4.3886,
314
+ "step": 5100
315
+ },
316
+ {
317
+ "epoch": 4.06,
318
+ "learning_rate": 0.0001,
319
+ "loss": 4.3915,
320
+ "step": 5200
321
+ },
322
+ {
323
+ "epoch": 4.14,
324
+ "learning_rate": 0.0001,
325
+ "loss": 4.3706,
326
+ "step": 5300
327
+ },
328
+ {
329
+ "epoch": 4.22,
330
+ "learning_rate": 0.0001,
331
+ "loss": 4.3528,
332
+ "step": 5400
333
+ },
334
+ {
335
+ "epoch": 4.3,
336
+ "learning_rate": 0.0001,
337
+ "loss": 4.3272,
338
+ "step": 5500
339
+ },
340
+ {
341
+ "epoch": 4.38,
342
+ "learning_rate": 0.0001,
343
+ "loss": 4.3679,
344
+ "step": 5600
345
+ },
346
+ {
347
+ "epoch": 4.45,
348
+ "learning_rate": 0.0001,
349
+ "loss": 4.3608,
350
+ "step": 5700
351
+ },
352
+ {
353
+ "epoch": 4.53,
354
+ "learning_rate": 0.0001,
355
+ "loss": 4.3262,
356
+ "step": 5800
357
+ },
358
+ {
359
+ "epoch": 4.61,
360
+ "learning_rate": 0.0001,
361
+ "loss": 4.3097,
362
+ "step": 5900
363
+ },
364
+ {
365
+ "epoch": 4.69,
366
+ "learning_rate": 0.0001,
367
+ "loss": 4.2973,
368
+ "step": 6000
369
+ },
370
+ {
371
+ "epoch": 4.77,
372
+ "learning_rate": 0.0001,
373
+ "loss": 4.2756,
374
+ "step": 6100
375
+ },
376
+ {
377
+ "epoch": 4.84,
378
+ "learning_rate": 0.0001,
379
+ "loss": 4.3174,
380
+ "step": 6200
381
+ },
382
+ {
383
+ "epoch": 4.92,
384
+ "learning_rate": 0.0001,
385
+ "loss": 4.2786,
386
+ "step": 6300
387
+ },
388
+ {
389
+ "epoch": 5.0,
390
+ "learning_rate": 0.0001,
391
+ "loss": 4.2484,
392
+ "step": 6400
393
+ },
394
+ {
395
+ "epoch": 5.08,
396
+ "learning_rate": 0.0001,
397
+ "loss": 4.2634,
398
+ "step": 6500
399
+ },
400
+ {
401
+ "epoch": 5.16,
402
+ "learning_rate": 0.0001,
403
+ "loss": 4.228,
404
+ "step": 6600
405
+ },
406
+ {
407
+ "epoch": 5.23,
408
+ "learning_rate": 0.0001,
409
+ "loss": 4.2745,
410
+ "step": 6700
411
+ },
412
+ {
413
+ "epoch": 5.31,
414
+ "learning_rate": 0.0001,
415
+ "loss": 4.2297,
416
+ "step": 6800
417
+ },
418
+ {
419
+ "epoch": 5.39,
420
+ "learning_rate": 0.0001,
421
+ "loss": 4.1989,
422
+ "step": 6900
423
+ },
424
+ {
425
+ "epoch": 5.47,
426
+ "learning_rate": 0.0001,
427
+ "loss": 4.2074,
428
+ "step": 7000
429
+ },
430
+ {
431
+ "epoch": 5.55,
432
+ "learning_rate": 0.0001,
433
+ "loss": 4.201,
434
+ "step": 7100
435
+ },
436
+ {
437
+ "epoch": 5.62,
438
+ "learning_rate": 0.0001,
439
+ "loss": 4.2109,
440
+ "step": 7200
441
+ },
442
+ {
443
+ "epoch": 5.7,
444
+ "learning_rate": 0.0001,
445
+ "loss": 4.1945,
446
+ "step": 7300
447
+ },
448
+ {
449
+ "epoch": 5.78,
450
+ "learning_rate": 0.0001,
451
+ "loss": 4.1744,
452
+ "step": 7400
453
+ },
454
+ {
455
+ "epoch": 5.86,
456
+ "learning_rate": 0.0001,
457
+ "loss": 4.1741,
458
+ "step": 7500
459
+ },
460
+ {
461
+ "epoch": 5.94,
462
+ "learning_rate": 0.0001,
463
+ "loss": 4.19,
464
+ "step": 7600
465
+ },
466
+ {
467
+ "epoch": 6.02,
468
+ "learning_rate": 0.0001,
469
+ "loss": 4.1612,
470
+ "step": 7700
471
+ },
472
+ {
473
+ "epoch": 6.09,
474
+ "learning_rate": 0.0001,
475
+ "loss": 4.1559,
476
+ "step": 7800
477
+ },
478
+ {
479
+ "epoch": 6.17,
480
+ "learning_rate": 0.0001,
481
+ "loss": 4.1378,
482
+ "step": 7900
483
+ },
484
+ {
485
+ "epoch": 6.25,
486
+ "learning_rate": 0.0001,
487
+ "loss": 4.144,
488
+ "step": 8000
489
+ },
490
+ {
491
+ "epoch": 6.33,
492
+ "learning_rate": 0.0001,
493
+ "loss": 4.1601,
494
+ "step": 8100
495
+ },
496
+ {
497
+ "epoch": 6.41,
498
+ "learning_rate": 0.0001,
499
+ "loss": 4.1339,
500
+ "step": 8200
501
+ },
502
+ {
503
+ "epoch": 6.48,
504
+ "learning_rate": 0.0001,
505
+ "loss": 4.13,
506
+ "step": 8300
507
+ },
508
+ {
509
+ "epoch": 6.56,
510
+ "learning_rate": 0.0001,
511
+ "loss": 4.0877,
512
+ "step": 8400
513
+ },
514
+ {
515
+ "epoch": 6.64,
516
+ "learning_rate": 0.0001,
517
+ "loss": 4.1022,
518
+ "step": 8500
519
+ },
520
+ {
521
+ "epoch": 6.72,
522
+ "learning_rate": 0.0001,
523
+ "loss": 4.1034,
524
+ "step": 8600
525
+ },
526
+ {
527
+ "epoch": 6.8,
528
+ "learning_rate": 0.0001,
529
+ "loss": 4.091,
530
+ "step": 8700
531
+ },
532
+ {
533
+ "epoch": 6.88,
534
+ "learning_rate": 0.0001,
535
+ "loss": 4.0914,
536
+ "step": 8800
537
+ },
538
+ {
539
+ "epoch": 6.95,
540
+ "learning_rate": 0.0001,
541
+ "loss": 4.0937,
542
+ "step": 8900
543
+ },
544
+ {
545
+ "epoch": 7.03,
546
+ "learning_rate": 0.0001,
547
+ "loss": 4.0897,
548
+ "step": 9000
549
+ },
550
+ {
551
+ "epoch": 7.11,
552
+ "learning_rate": 0.0001,
553
+ "loss": 4.0844,
554
+ "step": 9100
555
+ },
556
+ {
557
+ "epoch": 7.19,
558
+ "learning_rate": 0.0001,
559
+ "loss": 4.0574,
560
+ "step": 9200
561
+ },
562
+ {
563
+ "epoch": 7.27,
564
+ "learning_rate": 0.0001,
565
+ "loss": 4.0837,
566
+ "step": 9300
567
+ },
568
+ {
569
+ "epoch": 7.34,
570
+ "learning_rate": 0.0001,
571
+ "loss": 4.0766,
572
+ "step": 9400
573
+ },
574
+ {
575
+ "epoch": 7.42,
576
+ "learning_rate": 0.0001,
577
+ "loss": 4.066,
578
+ "step": 9500
579
+ },
580
+ {
581
+ "epoch": 7.5,
582
+ "learning_rate": 0.0001,
583
+ "loss": 3.9944,
584
+ "step": 9600
585
+ },
586
+ {
587
+ "epoch": 7.58,
588
+ "learning_rate": 0.0001,
589
+ "loss": 4.0126,
590
+ "step": 9700
591
+ },
592
+ {
593
+ "epoch": 7.66,
594
+ "learning_rate": 0.0001,
595
+ "loss": 4.0273,
596
+ "step": 9800
597
+ },
598
+ {
599
+ "epoch": 7.73,
600
+ "learning_rate": 0.0001,
601
+ "loss": 4.0706,
602
+ "step": 9900
603
+ },
604
+ {
605
+ "epoch": 7.81,
606
+ "learning_rate": 0.0001,
607
+ "loss": 4.0008,
608
+ "step": 10000
609
+ },
610
+ {
611
+ "epoch": 7.89,
612
+ "learning_rate": 0.0001,
613
+ "loss": 4.0323,
614
+ "step": 10100
615
+ },
616
+ {
617
+ "epoch": 7.97,
618
+ "learning_rate": 0.0001,
619
+ "loss": 3.9864,
620
+ "step": 10200
621
+ },
622
+ {
623
+ "epoch": 8.05,
624
+ "learning_rate": 0.0001,
625
+ "loss": 4.0265,
626
+ "step": 10300
627
+ },
628
+ {
629
+ "epoch": 8.12,
630
+ "learning_rate": 0.0001,
631
+ "loss": 3.9874,
632
+ "step": 10400
633
+ },
634
+ {
635
+ "epoch": 8.2,
636
+ "learning_rate": 0.0001,
637
+ "loss": 3.9839,
638
+ "step": 10500
639
+ },
640
+ {
641
+ "epoch": 8.28,
642
+ "learning_rate": 0.0001,
643
+ "loss": 3.9583,
644
+ "step": 10600
645
+ },
646
+ {
647
+ "epoch": 8.36,
648
+ "learning_rate": 0.0001,
649
+ "loss": 4.007,
650
+ "step": 10700
651
+ },
652
+ {
653
+ "epoch": 8.44,
654
+ "learning_rate": 0.0001,
655
+ "loss": 3.9567,
656
+ "step": 10800
657
+ },
658
+ {
659
+ "epoch": 8.52,
660
+ "learning_rate": 0.0001,
661
+ "loss": 3.9888,
662
+ "step": 10900
663
+ },
664
+ {
665
+ "epoch": 8.59,
666
+ "learning_rate": 0.0001,
667
+ "loss": 3.9993,
668
+ "step": 11000
669
+ },
670
+ {
671
+ "epoch": 8.67,
672
+ "learning_rate": 0.0001,
673
+ "loss": 3.9914,
674
+ "step": 11100
675
+ },
676
+ {
677
+ "epoch": 8.75,
678
+ "learning_rate": 0.0001,
679
+ "loss": 4.0094,
680
+ "step": 11200
681
+ },
682
+ {
683
+ "epoch": 8.83,
684
+ "learning_rate": 0.0001,
685
+ "loss": 3.9338,
686
+ "step": 11300
687
+ },
688
+ {
689
+ "epoch": 8.91,
690
+ "learning_rate": 0.0001,
691
+ "loss": 3.9472,
692
+ "step": 11400
693
+ },
694
+ {
695
+ "epoch": 8.98,
696
+ "learning_rate": 0.0001,
697
+ "loss": 3.9595,
698
+ "step": 11500
699
+ },
700
+ {
701
+ "epoch": 9.06,
702
+ "learning_rate": 0.0001,
703
+ "loss": 3.9545,
704
+ "step": 11600
705
+ },
706
+ {
707
+ "epoch": 9.14,
708
+ "learning_rate": 0.0001,
709
+ "loss": 3.993,
710
+ "step": 11700
711
+ },
712
+ {
713
+ "epoch": 9.22,
714
+ "learning_rate": 0.0001,
715
+ "loss": 3.895,
716
+ "step": 11800
717
+ },
718
+ {
719
+ "epoch": 9.3,
720
+ "learning_rate": 0.0001,
721
+ "loss": 3.9527,
722
+ "step": 11900
723
+ },
724
+ {
725
+ "epoch": 9.38,
726
+ "learning_rate": 0.0001,
727
+ "loss": 3.9205,
728
+ "step": 12000
729
+ },
730
+ {
731
+ "epoch": 9.45,
732
+ "learning_rate": 0.0001,
733
+ "loss": 3.914,
734
+ "step": 12100
735
+ },
736
+ {
737
+ "epoch": 9.53,
738
+ "learning_rate": 0.0001,
739
+ "loss": 3.8872,
740
+ "step": 12200
741
+ },
742
+ {
743
+ "epoch": 9.61,
744
+ "learning_rate": 0.0001,
745
+ "loss": 3.9171,
746
+ "step": 12300
747
+ }
748
+ ],
749
+ "max_steps": 12800,
750
+ "num_train_epochs": 10,
751
+ "total_flos": 9246417995335680.0,
752
+ "trial_name": null,
753
+ "trial_params": null
754
+ }
checkpoint-12360/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e683e2614c4dcf1ee5d4bfe9622ed2cfdbf3faf9319855e4a9b393368f21e44
3
+ size 3899
checkpoint-12390/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: fp4
15
+ - bnb_4bit_use_double_quant: False
16
+ - bnb_4bit_compute_dtype: float32
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.5.0.dev0
checkpoint-12390/adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "models/TinyStories-33M",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 8,
11
+ "lora_dropout": 0.0,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 16,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
checkpoint-12390/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2638e3a4999749d4f6a329f2d948a026f296cc0606de56f66eb59a7a5435ae82
3
+ size 792353
checkpoint-12390/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:611d9b387269a4b7f63640b47040166c4f161581bf62376c80b680134454b7c6
3
+ size 1586873
checkpoint-12390/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8bb3860d98363e44bc4704fb9458deec927fd96997d1d09c80639070db959d22
3
+ size 14575
checkpoint-12390/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd03d1ae0c09c9520cd92e5bf2331a87151cb6edff37a2a3de420ef62b3a176f
3
+ size 627
checkpoint-12390/trainer_state.json ADDED
@@ -0,0 +1,754 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 9.6796875,
5
+ "global_step": 12390,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.08,
12
+ "learning_rate": 0.0001,
13
+ "loss": 10.386,
14
+ "step": 100
15
+ },
16
+ {
17
+ "epoch": 0.16,
18
+ "learning_rate": 0.0001,
19
+ "loss": 7.4061,
20
+ "step": 200
21
+ },
22
+ {
23
+ "epoch": 0.23,
24
+ "learning_rate": 0.0001,
25
+ "loss": 6.4112,
26
+ "step": 300
27
+ },
28
+ {
29
+ "epoch": 0.31,
30
+ "learning_rate": 0.0001,
31
+ "loss": 6.091,
32
+ "step": 400
33
+ },
34
+ {
35
+ "epoch": 0.39,
36
+ "learning_rate": 0.0001,
37
+ "loss": 5.8571,
38
+ "step": 500
39
+ },
40
+ {
41
+ "epoch": 0.47,
42
+ "learning_rate": 0.0001,
43
+ "loss": 5.7144,
44
+ "step": 600
45
+ },
46
+ {
47
+ "epoch": 0.55,
48
+ "learning_rate": 0.0001,
49
+ "loss": 5.6228,
50
+ "step": 700
51
+ },
52
+ {
53
+ "epoch": 0.62,
54
+ "learning_rate": 0.0001,
55
+ "loss": 5.546,
56
+ "step": 800
57
+ },
58
+ {
59
+ "epoch": 0.7,
60
+ "learning_rate": 0.0001,
61
+ "loss": 5.4904,
62
+ "step": 900
63
+ },
64
+ {
65
+ "epoch": 0.78,
66
+ "learning_rate": 0.0001,
67
+ "loss": 5.4413,
68
+ "step": 1000
69
+ },
70
+ {
71
+ "epoch": 0.86,
72
+ "learning_rate": 0.0001,
73
+ "loss": 5.3603,
74
+ "step": 1100
75
+ },
76
+ {
77
+ "epoch": 0.94,
78
+ "learning_rate": 0.0001,
79
+ "loss": 5.3346,
80
+ "step": 1200
81
+ },
82
+ {
83
+ "epoch": 1.02,
84
+ "learning_rate": 0.0001,
85
+ "loss": 5.2744,
86
+ "step": 1300
87
+ },
88
+ {
89
+ "epoch": 1.09,
90
+ "learning_rate": 0.0001,
91
+ "loss": 5.231,
92
+ "step": 1400
93
+ },
94
+ {
95
+ "epoch": 1.17,
96
+ "learning_rate": 0.0001,
97
+ "loss": 5.1782,
98
+ "step": 1500
99
+ },
100
+ {
101
+ "epoch": 1.25,
102
+ "learning_rate": 0.0001,
103
+ "loss": 5.1613,
104
+ "step": 1600
105
+ },
106
+ {
107
+ "epoch": 1.33,
108
+ "learning_rate": 0.0001,
109
+ "loss": 5.1091,
110
+ "step": 1700
111
+ },
112
+ {
113
+ "epoch": 1.41,
114
+ "learning_rate": 0.0001,
115
+ "loss": 5.0939,
116
+ "step": 1800
117
+ },
118
+ {
119
+ "epoch": 1.48,
120
+ "learning_rate": 0.0001,
121
+ "loss": 5.0667,
122
+ "step": 1900
123
+ },
124
+ {
125
+ "epoch": 1.56,
126
+ "learning_rate": 0.0001,
127
+ "loss": 5.0054,
128
+ "step": 2000
129
+ },
130
+ {
131
+ "epoch": 1.64,
132
+ "learning_rate": 0.0001,
133
+ "loss": 4.96,
134
+ "step": 2100
135
+ },
136
+ {
137
+ "epoch": 1.72,
138
+ "learning_rate": 0.0001,
139
+ "loss": 4.9301,
140
+ "step": 2200
141
+ },
142
+ {
143
+ "epoch": 1.8,
144
+ "learning_rate": 0.0001,
145
+ "loss": 4.9302,
146
+ "step": 2300
147
+ },
148
+ {
149
+ "epoch": 1.88,
150
+ "learning_rate": 0.0001,
151
+ "loss": 4.9129,
152
+ "step": 2400
153
+ },
154
+ {
155
+ "epoch": 1.95,
156
+ "learning_rate": 0.0001,
157
+ "loss": 4.8627,
158
+ "step": 2500
159
+ },
160
+ {
161
+ "epoch": 2.03,
162
+ "learning_rate": 0.0001,
163
+ "loss": 4.8336,
164
+ "step": 2600
165
+ },
166
+ {
167
+ "epoch": 2.11,
168
+ "learning_rate": 0.0001,
169
+ "loss": 4.8018,
170
+ "step": 2700
171
+ },
172
+ {
173
+ "epoch": 2.19,
174
+ "learning_rate": 0.0001,
175
+ "loss": 4.7479,
176
+ "step": 2800
177
+ },
178
+ {
179
+ "epoch": 2.27,
180
+ "learning_rate": 0.0001,
181
+ "loss": 4.7572,
182
+ "step": 2900
183
+ },
184
+ {
185
+ "epoch": 2.34,
186
+ "learning_rate": 0.0001,
187
+ "loss": 4.759,
188
+ "step": 3000
189
+ },
190
+ {
191
+ "epoch": 2.42,
192
+ "learning_rate": 0.0001,
193
+ "loss": 4.7441,
194
+ "step": 3100
195
+ },
196
+ {
197
+ "epoch": 2.5,
198
+ "learning_rate": 0.0001,
199
+ "loss": 4.7195,
200
+ "step": 3200
201
+ },
202
+ {
203
+ "epoch": 2.58,
204
+ "learning_rate": 0.0001,
205
+ "loss": 4.6821,
206
+ "step": 3300
207
+ },
208
+ {
209
+ "epoch": 2.66,
210
+ "learning_rate": 0.0001,
211
+ "loss": 4.6587,
212
+ "step": 3400
213
+ },
214
+ {
215
+ "epoch": 2.73,
216
+ "learning_rate": 0.0001,
217
+ "loss": 4.6228,
218
+ "step": 3500
219
+ },
220
+ {
221
+ "epoch": 2.81,
222
+ "learning_rate": 0.0001,
223
+ "loss": 4.6248,
224
+ "step": 3600
225
+ },
226
+ {
227
+ "epoch": 2.89,
228
+ "learning_rate": 0.0001,
229
+ "loss": 4.5801,
230
+ "step": 3700
231
+ },
232
+ {
233
+ "epoch": 2.97,
234
+ "learning_rate": 0.0001,
235
+ "loss": 4.5934,
236
+ "step": 3800
237
+ },
238
+ {
239
+ "epoch": 3.05,
240
+ "learning_rate": 0.0001,
241
+ "loss": 4.5781,
242
+ "step": 3900
243
+ },
244
+ {
245
+ "epoch": 3.12,
246
+ "learning_rate": 0.0001,
247
+ "loss": 4.5763,
248
+ "step": 4000
249
+ },
250
+ {
251
+ "epoch": 3.2,
252
+ "learning_rate": 0.0001,
253
+ "loss": 4.5598,
254
+ "step": 4100
255
+ },
256
+ {
257
+ "epoch": 3.28,
258
+ "learning_rate": 0.0001,
259
+ "loss": 4.4796,
260
+ "step": 4200
261
+ },
262
+ {
263
+ "epoch": 3.36,
264
+ "learning_rate": 0.0001,
265
+ "loss": 4.5204,
266
+ "step": 4300
267
+ },
268
+ {
269
+ "epoch": 3.44,
270
+ "learning_rate": 0.0001,
271
+ "loss": 4.4858,
272
+ "step": 4400
273
+ },
274
+ {
275
+ "epoch": 3.52,
276
+ "learning_rate": 0.0001,
277
+ "loss": 4.4882,
278
+ "step": 4500
279
+ },
280
+ {
281
+ "epoch": 3.59,
282
+ "learning_rate": 0.0001,
283
+ "loss": 4.4852,
284
+ "step": 4600
285
+ },
286
+ {
287
+ "epoch": 3.67,
288
+ "learning_rate": 0.0001,
289
+ "loss": 4.4302,
290
+ "step": 4700
291
+ },
292
+ {
293
+ "epoch": 3.75,
294
+ "learning_rate": 0.0001,
295
+ "loss": 4.4267,
296
+ "step": 4800
297
+ },
298
+ {
299
+ "epoch": 3.83,
300
+ "learning_rate": 0.0001,
301
+ "loss": 4.4298,
302
+ "step": 4900
303
+ },
304
+ {
305
+ "epoch": 3.91,
306
+ "learning_rate": 0.0001,
307
+ "loss": 4.3892,
308
+ "step": 5000
309
+ },
310
+ {
311
+ "epoch": 3.98,
312
+ "learning_rate": 0.0001,
313
+ "loss": 4.3886,
314
+ "step": 5100
315
+ },
316
+ {
317
+ "epoch": 4.06,
318
+ "learning_rate": 0.0001,
319
+ "loss": 4.3915,
320
+ "step": 5200
321
+ },
322
+ {
323
+ "epoch": 4.14,
324
+ "learning_rate": 0.0001,
325
+ "loss": 4.3706,
326
+ "step": 5300
327
+ },
328
+ {
329
+ "epoch": 4.22,
330
+ "learning_rate": 0.0001,
331
+ "loss": 4.3528,
332
+ "step": 5400
333
+ },
334
+ {
335
+ "epoch": 4.3,
336
+ "learning_rate": 0.0001,
337
+ "loss": 4.3272,
338
+ "step": 5500
339
+ },
340
+ {
341
+ "epoch": 4.38,
342
+ "learning_rate": 0.0001,
343
+ "loss": 4.3679,
344
+ "step": 5600
345
+ },
346
+ {
347
+ "epoch": 4.45,
348
+ "learning_rate": 0.0001,
349
+ "loss": 4.3608,
350
+ "step": 5700
351
+ },
352
+ {
353
+ "epoch": 4.53,
354
+ "learning_rate": 0.0001,
355
+ "loss": 4.3262,
356
+ "step": 5800
357
+ },
358
+ {
359
+ "epoch": 4.61,
360
+ "learning_rate": 0.0001,
361
+ "loss": 4.3097,
362
+ "step": 5900
363
+ },
364
+ {
365
+ "epoch": 4.69,
366
+ "learning_rate": 0.0001,
367
+ "loss": 4.2973,
368
+ "step": 6000
369
+ },
370
+ {
371
+ "epoch": 4.77,
372
+ "learning_rate": 0.0001,
373
+ "loss": 4.2756,
374
+ "step": 6100
375
+ },
376
+ {
377
+ "epoch": 4.84,
378
+ "learning_rate": 0.0001,
379
+ "loss": 4.3174,
380
+ "step": 6200
381
+ },
382
+ {
383
+ "epoch": 4.92,
384
+ "learning_rate": 0.0001,
385
+ "loss": 4.2786,
386
+ "step": 6300
387
+ },
388
+ {
389
+ "epoch": 5.0,
390
+ "learning_rate": 0.0001,
391
+ "loss": 4.2484,
392
+ "step": 6400
393
+ },
394
+ {
395
+ "epoch": 5.08,
396
+ "learning_rate": 0.0001,
397
+ "loss": 4.2634,
398
+ "step": 6500
399
+ },
400
+ {
401
+ "epoch": 5.16,
402
+ "learning_rate": 0.0001,
403
+ "loss": 4.228,
404
+ "step": 6600
405
+ },
406
+ {
407
+ "epoch": 5.23,
408
+ "learning_rate": 0.0001,
409
+ "loss": 4.2745,
410
+ "step": 6700
411
+ },
412
+ {
413
+ "epoch": 5.31,
414
+ "learning_rate": 0.0001,
415
+ "loss": 4.2297,
416
+ "step": 6800
417
+ },
418
+ {
419
+ "epoch": 5.39,
420
+ "learning_rate": 0.0001,
421
+ "loss": 4.1989,
422
+ "step": 6900
423
+ },
424
+ {
425
+ "epoch": 5.47,
426
+ "learning_rate": 0.0001,
427
+ "loss": 4.2074,
428
+ "step": 7000
429
+ },
430
+ {
431
+ "epoch": 5.55,
432
+ "learning_rate": 0.0001,
433
+ "loss": 4.201,
434
+ "step": 7100
435
+ },
436
+ {
437
+ "epoch": 5.62,
438
+ "learning_rate": 0.0001,
439
+ "loss": 4.2109,
440
+ "step": 7200
441
+ },
442
+ {
443
+ "epoch": 5.7,
444
+ "learning_rate": 0.0001,
445
+ "loss": 4.1945,
446
+ "step": 7300
447
+ },
448
+ {
449
+ "epoch": 5.78,
450
+ "learning_rate": 0.0001,
451
+ "loss": 4.1744,
452
+ "step": 7400
453
+ },
454
+ {
455
+ "epoch": 5.86,
456
+ "learning_rate": 0.0001,
457
+ "loss": 4.1741,
458
+ "step": 7500
459
+ },
460
+ {
461
+ "epoch": 5.94,
462
+ "learning_rate": 0.0001,
463
+ "loss": 4.19,
464
+ "step": 7600
465
+ },
466
+ {
467
+ "epoch": 6.02,
468
+ "learning_rate": 0.0001,
469
+ "loss": 4.1612,
470
+ "step": 7700
471
+ },
472
+ {
473
+ "epoch": 6.09,
474
+ "learning_rate": 0.0001,
475
+ "loss": 4.1559,
476
+ "step": 7800
477
+ },
478
+ {
479
+ "epoch": 6.17,
480
+ "learning_rate": 0.0001,
481
+ "loss": 4.1378,
482
+ "step": 7900
483
+ },
484
+ {
485
+ "epoch": 6.25,
486
+ "learning_rate": 0.0001,
487
+ "loss": 4.144,
488
+ "step": 8000
489
+ },
490
+ {
491
+ "epoch": 6.33,
492
+ "learning_rate": 0.0001,
493
+ "loss": 4.1601,
494
+ "step": 8100
495
+ },
496
+ {
497
+ "epoch": 6.41,
498
+ "learning_rate": 0.0001,
499
+ "loss": 4.1339,
500
+ "step": 8200
501
+ },
502
+ {
503
+ "epoch": 6.48,
504
+ "learning_rate": 0.0001,
505
+ "loss": 4.13,
506
+ "step": 8300
507
+ },
508
+ {
509
+ "epoch": 6.56,
510
+ "learning_rate": 0.0001,
511
+ "loss": 4.0877,
512
+ "step": 8400
513
+ },
514
+ {
515
+ "epoch": 6.64,
516
+ "learning_rate": 0.0001,
517
+ "loss": 4.1022,
518
+ "step": 8500
519
+ },
520
+ {
521
+ "epoch": 6.72,
522
+ "learning_rate": 0.0001,
523
+ "loss": 4.1034,
524
+ "step": 8600
525
+ },
526
+ {
527
+ "epoch": 6.8,
528
+ "learning_rate": 0.0001,
529
+ "loss": 4.091,
530
+ "step": 8700
531
+ },
532
+ {
533
+ "epoch": 6.88,
534
+ "learning_rate": 0.0001,
535
+ "loss": 4.0914,
536
+ "step": 8800
537
+ },
538
+ {
539
+ "epoch": 6.95,
540
+ "learning_rate": 0.0001,
541
+ "loss": 4.0937,
542
+ "step": 8900
543
+ },
544
+ {
545
+ "epoch": 7.03,
546
+ "learning_rate": 0.0001,
547
+ "loss": 4.0897,
548
+ "step": 9000
549
+ },
550
+ {
551
+ "epoch": 7.11,
552
+ "learning_rate": 0.0001,
553
+ "loss": 4.0844,
554
+ "step": 9100
555
+ },
556
+ {
557
+ "epoch": 7.19,
558
+ "learning_rate": 0.0001,
559
+ "loss": 4.0574,
560
+ "step": 9200
561
+ },
562
+ {
563
+ "epoch": 7.27,
564
+ "learning_rate": 0.0001,
565
+ "loss": 4.0837,
566
+ "step": 9300
567
+ },
568
+ {
569
+ "epoch": 7.34,
570
+ "learning_rate": 0.0001,
571
+ "loss": 4.0766,
572
+ "step": 9400
573
+ },
574
+ {
575
+ "epoch": 7.42,
576
+ "learning_rate": 0.0001,
577
+ "loss": 4.066,
578
+ "step": 9500
579
+ },
580
+ {
581
+ "epoch": 7.5,
582
+ "learning_rate": 0.0001,
583
+ "loss": 3.9944,
584
+ "step": 9600
585
+ },
586
+ {
587
+ "epoch": 7.58,
588
+ "learning_rate": 0.0001,
589
+ "loss": 4.0126,
590
+ "step": 9700
591
+ },
592
+ {
593
+ "epoch": 7.66,
594
+ "learning_rate": 0.0001,
595
+ "loss": 4.0273,
596
+ "step": 9800
597
+ },
598
+ {
599
+ "epoch": 7.73,
600
+ "learning_rate": 0.0001,
601
+ "loss": 4.0706,
602
+ "step": 9900
603
+ },
604
+ {
605
+ "epoch": 7.81,
606
+ "learning_rate": 0.0001,
607
+ "loss": 4.0008,
608
+ "step": 10000
609
+ },
610
+ {
611
+ "epoch": 7.89,
612
+ "learning_rate": 0.0001,
613
+ "loss": 4.0323,
614
+ "step": 10100
615
+ },
616
+ {
617
+ "epoch": 7.97,
618
+ "learning_rate": 0.0001,
619
+ "loss": 3.9864,
620
+ "step": 10200
621
+ },
622
+ {
623
+ "epoch": 8.05,
624
+ "learning_rate": 0.0001,
625
+ "loss": 4.0265,
626
+ "step": 10300
627
+ },
628
+ {
629
+ "epoch": 8.12,
630
+ "learning_rate": 0.0001,
631
+ "loss": 3.9874,
632
+ "step": 10400
633
+ },
634
+ {
635
+ "epoch": 8.2,
636
+ "learning_rate": 0.0001,
637
+ "loss": 3.9839,
638
+ "step": 10500
639
+ },
640
+ {
641
+ "epoch": 8.28,
642
+ "learning_rate": 0.0001,
643
+ "loss": 3.9583,
644
+ "step": 10600
645
+ },
646
+ {
647
+ "epoch": 8.36,
648
+ "learning_rate": 0.0001,
649
+ "loss": 4.007,
650
+ "step": 10700
651
+ },
652
+ {
653
+ "epoch": 8.44,
654
+ "learning_rate": 0.0001,
655
+ "loss": 3.9567,
656
+ "step": 10800
657
+ },
658
+ {
659
+ "epoch": 8.52,
660
+ "learning_rate": 0.0001,
661
+ "loss": 3.9888,
662
+ "step": 10900
663
+ },
664
+ {
665
+ "epoch": 8.59,
666
+ "learning_rate": 0.0001,
667
+ "loss": 3.9993,
668
+ "step": 11000
669
+ },
670
+ {
671
+ "epoch": 8.67,
672
+ "learning_rate": 0.0001,
673
+ "loss": 3.9914,
674
+ "step": 11100
675
+ },
676
+ {
677
+ "epoch": 8.75,
678
+ "learning_rate": 0.0001,
679
+ "loss": 4.0094,
680
+ "step": 11200
681
+ },
682
+ {
683
+ "epoch": 8.83,
684
+ "learning_rate": 0.0001,
685
+ "loss": 3.9338,
686
+ "step": 11300
687
+ },
688
+ {
689
+ "epoch": 8.91,
690
+ "learning_rate": 0.0001,
691
+ "loss": 3.9472,
692
+ "step": 11400
693
+ },
694
+ {
695
+ "epoch": 8.98,
696
+ "learning_rate": 0.0001,
697
+ "loss": 3.9595,
698
+ "step": 11500
699
+ },
700
+ {
701
+ "epoch": 9.06,
702
+ "learning_rate": 0.0001,
703
+ "loss": 3.9545,
704
+ "step": 11600
705
+ },
706
+ {
707
+ "epoch": 9.14,
708
+ "learning_rate": 0.0001,
709
+ "loss": 3.993,
710
+ "step": 11700
711
+ },
712
+ {
713
+ "epoch": 9.22,
714
+ "learning_rate": 0.0001,
715
+ "loss": 3.895,
716
+ "step": 11800
717
+ },
718
+ {
719
+ "epoch": 9.3,
720
+ "learning_rate": 0.0001,
721
+ "loss": 3.9527,
722
+ "step": 11900
723
+ },
724
+ {
725
+ "epoch": 9.38,
726
+ "learning_rate": 0.0001,
727
+ "loss": 3.9205,
728
+ "step": 12000
729
+ },
730
+ {
731
+ "epoch": 9.45,
732
+ "learning_rate": 0.0001,
733
+ "loss": 3.914,
734
+ "step": 12100
735
+ },
736
+ {
737
+ "epoch": 9.53,
738
+ "learning_rate": 0.0001,
739
+ "loss": 3.8872,
740
+ "step": 12200
741
+ },
742
+ {
743
+ "epoch": 9.61,
744
+ "learning_rate": 0.0001,
745
+ "loss": 3.9171,
746
+ "step": 12300
747
+ }
748
+ ],
749
+ "max_steps": 12800,
750
+ "num_train_epochs": 10,
751
+ "total_flos": 9270435050588160.0,
752
+ "trial_name": null,
753
+ "trial_params": null
754
+ }
checkpoint-12390/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e683e2614c4dcf1ee5d4bfe9622ed2cfdbf3faf9319855e4a9b393368f21e44
3
+ size 3899
checkpoint-12420/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: fp4
15
+ - bnb_4bit_use_double_quant: False
16
+ - bnb_4bit_compute_dtype: float32
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.5.0.dev0
checkpoint-12420/adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "models/TinyStories-33M",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 8,
11
+ "lora_dropout": 0.0,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 16,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
checkpoint-12420/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:209158f1892bccba2617213b558a39d11cddd1e5136016bb669fc187b741f152
3
+ size 792353
checkpoint-12420/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:306b66f63a98bc0870cc880ebaf5b767253138b2e0f8e2badd78119ae6ed0284
3
+ size 1586873
checkpoint-12420/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8bb3860d98363e44bc4704fb9458deec927fd96997d1d09c80639070db959d22
3
+ size 14575
checkpoint-12420/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:67effef7e1527b34cecc0682500e658ff0a9d2cce8252039f87084b77aabfa5d
3
+ size 627
checkpoint-12420/trainer_state.json ADDED
@@ -0,0 +1,760 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 9.703125,
5
+ "global_step": 12420,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.08,
12
+ "learning_rate": 0.0001,
13
+ "loss": 10.386,
14
+ "step": 100
15
+ },
16
+ {
17
+ "epoch": 0.16,
18
+ "learning_rate": 0.0001,
19
+ "loss": 7.4061,
20
+ "step": 200
21
+ },
22
+ {
23
+ "epoch": 0.23,
24
+ "learning_rate": 0.0001,
25
+ "loss": 6.4112,
26
+ "step": 300
27
+ },
28
+ {
29
+ "epoch": 0.31,
30
+ "learning_rate": 0.0001,
31
+ "loss": 6.091,
32
+ "step": 400
33
+ },
34
+ {
35
+ "epoch": 0.39,
36
+ "learning_rate": 0.0001,
37
+ "loss": 5.8571,
38
+ "step": 500
39
+ },
40
+ {
41
+ "epoch": 0.47,
42
+ "learning_rate": 0.0001,
43
+ "loss": 5.7144,
44
+ "step": 600
45
+ },
46
+ {
47
+ "epoch": 0.55,
48
+ "learning_rate": 0.0001,
49
+ "loss": 5.6228,
50
+ "step": 700
51
+ },
52
+ {
53
+ "epoch": 0.62,
54
+ "learning_rate": 0.0001,
55
+ "loss": 5.546,
56
+ "step": 800
57
+ },
58
+ {
59
+ "epoch": 0.7,
60
+ "learning_rate": 0.0001,
61
+ "loss": 5.4904,
62
+ "step": 900
63
+ },
64
+ {
65
+ "epoch": 0.78,
66
+ "learning_rate": 0.0001,
67
+ "loss": 5.4413,
68
+ "step": 1000
69
+ },
70
+ {
71
+ "epoch": 0.86,
72
+ "learning_rate": 0.0001,
73
+ "loss": 5.3603,
74
+ "step": 1100
75
+ },
76
+ {
77
+ "epoch": 0.94,
78
+ "learning_rate": 0.0001,
79
+ "loss": 5.3346,
80
+ "step": 1200
81
+ },
82
+ {
83
+ "epoch": 1.02,
84
+ "learning_rate": 0.0001,
85
+ "loss": 5.2744,
86
+ "step": 1300
87
+ },
88
+ {
89
+ "epoch": 1.09,
90
+ "learning_rate": 0.0001,
91
+ "loss": 5.231,
92
+ "step": 1400
93
+ },
94
+ {
95
+ "epoch": 1.17,
96
+ "learning_rate": 0.0001,
97
+ "loss": 5.1782,
98
+ "step": 1500
99
+ },
100
+ {
101
+ "epoch": 1.25,
102
+ "learning_rate": 0.0001,
103
+ "loss": 5.1613,
104
+ "step": 1600
105
+ },
106
+ {
107
+ "epoch": 1.33,
108
+ "learning_rate": 0.0001,
109
+ "loss": 5.1091,
110
+ "step": 1700
111
+ },
112
+ {
113
+ "epoch": 1.41,
114
+ "learning_rate": 0.0001,
115
+ "loss": 5.0939,
116
+ "step": 1800
117
+ },
118
+ {
119
+ "epoch": 1.48,
120
+ "learning_rate": 0.0001,
121
+ "loss": 5.0667,
122
+ "step": 1900
123
+ },
124
+ {
125
+ "epoch": 1.56,
126
+ "learning_rate": 0.0001,
127
+ "loss": 5.0054,
128
+ "step": 2000
129
+ },
130
+ {
131
+ "epoch": 1.64,
132
+ "learning_rate": 0.0001,
133
+ "loss": 4.96,
134
+ "step": 2100
135
+ },
136
+ {
137
+ "epoch": 1.72,
138
+ "learning_rate": 0.0001,
139
+ "loss": 4.9301,
140
+ "step": 2200
141
+ },
142
+ {
143
+ "epoch": 1.8,
144
+ "learning_rate": 0.0001,
145
+ "loss": 4.9302,
146
+ "step": 2300
147
+ },
148
+ {
149
+ "epoch": 1.88,
150
+ "learning_rate": 0.0001,
151
+ "loss": 4.9129,
152
+ "step": 2400
153
+ },
154
+ {
155
+ "epoch": 1.95,
156
+ "learning_rate": 0.0001,
157
+ "loss": 4.8627,
158
+ "step": 2500
159
+ },
160
+ {
161
+ "epoch": 2.03,
162
+ "learning_rate": 0.0001,
163
+ "loss": 4.8336,
164
+ "step": 2600
165
+ },
166
+ {
167
+ "epoch": 2.11,
168
+ "learning_rate": 0.0001,
169
+ "loss": 4.8018,
170
+ "step": 2700
171
+ },
172
+ {
173
+ "epoch": 2.19,
174
+ "learning_rate": 0.0001,
175
+ "loss": 4.7479,
176
+ "step": 2800
177
+ },
178
+ {
179
+ "epoch": 2.27,
180
+ "learning_rate": 0.0001,
181
+ "loss": 4.7572,
182
+ "step": 2900
183
+ },
184
+ {
185
+ "epoch": 2.34,
186
+ "learning_rate": 0.0001,
187
+ "loss": 4.759,
188
+ "step": 3000
189
+ },
190
+ {
191
+ "epoch": 2.42,
192
+ "learning_rate": 0.0001,
193
+ "loss": 4.7441,
194
+ "step": 3100
195
+ },
196
+ {
197
+ "epoch": 2.5,
198
+ "learning_rate": 0.0001,
199
+ "loss": 4.7195,
200
+ "step": 3200
201
+ },
202
+ {
203
+ "epoch": 2.58,
204
+ "learning_rate": 0.0001,
205
+ "loss": 4.6821,
206
+ "step": 3300
207
+ },
208
+ {
209
+ "epoch": 2.66,
210
+ "learning_rate": 0.0001,
211
+ "loss": 4.6587,
212
+ "step": 3400
213
+ },
214
+ {
215
+ "epoch": 2.73,
216
+ "learning_rate": 0.0001,
217
+ "loss": 4.6228,
218
+ "step": 3500
219
+ },
220
+ {
221
+ "epoch": 2.81,
222
+ "learning_rate": 0.0001,
223
+ "loss": 4.6248,
224
+ "step": 3600
225
+ },
226
+ {
227
+ "epoch": 2.89,
228
+ "learning_rate": 0.0001,
229
+ "loss": 4.5801,
230
+ "step": 3700
231
+ },
232
+ {
233
+ "epoch": 2.97,
234
+ "learning_rate": 0.0001,
235
+ "loss": 4.5934,
236
+ "step": 3800
237
+ },
238
+ {
239
+ "epoch": 3.05,
240
+ "learning_rate": 0.0001,
241
+ "loss": 4.5781,
242
+ "step": 3900
243
+ },
244
+ {
245
+ "epoch": 3.12,
246
+ "learning_rate": 0.0001,
247
+ "loss": 4.5763,
248
+ "step": 4000
249
+ },
250
+ {
251
+ "epoch": 3.2,
252
+ "learning_rate": 0.0001,
253
+ "loss": 4.5598,
254
+ "step": 4100
255
+ },
256
+ {
257
+ "epoch": 3.28,
258
+ "learning_rate": 0.0001,
259
+ "loss": 4.4796,
260
+ "step": 4200
261
+ },
262
+ {
263
+ "epoch": 3.36,
264
+ "learning_rate": 0.0001,
265
+ "loss": 4.5204,
266
+ "step": 4300
267
+ },
268
+ {
269
+ "epoch": 3.44,
270
+ "learning_rate": 0.0001,
271
+ "loss": 4.4858,
272
+ "step": 4400
273
+ },
274
+ {
275
+ "epoch": 3.52,
276
+ "learning_rate": 0.0001,
277
+ "loss": 4.4882,
278
+ "step": 4500
279
+ },
280
+ {
281
+ "epoch": 3.59,
282
+ "learning_rate": 0.0001,
283
+ "loss": 4.4852,
284
+ "step": 4600
285
+ },
286
+ {
287
+ "epoch": 3.67,
288
+ "learning_rate": 0.0001,
289
+ "loss": 4.4302,
290
+ "step": 4700
291
+ },
292
+ {
293
+ "epoch": 3.75,
294
+ "learning_rate": 0.0001,
295
+ "loss": 4.4267,
296
+ "step": 4800
297
+ },
298
+ {
299
+ "epoch": 3.83,
300
+ "learning_rate": 0.0001,
301
+ "loss": 4.4298,
302
+ "step": 4900
303
+ },
304
+ {
305
+ "epoch": 3.91,
306
+ "learning_rate": 0.0001,
307
+ "loss": 4.3892,
308
+ "step": 5000
309
+ },
310
+ {
311
+ "epoch": 3.98,
312
+ "learning_rate": 0.0001,
313
+ "loss": 4.3886,
314
+ "step": 5100
315
+ },
316
+ {
317
+ "epoch": 4.06,
318
+ "learning_rate": 0.0001,
319
+ "loss": 4.3915,
320
+ "step": 5200
321
+ },
322
+ {
323
+ "epoch": 4.14,
324
+ "learning_rate": 0.0001,
325
+ "loss": 4.3706,
326
+ "step": 5300
327
+ },
328
+ {
329
+ "epoch": 4.22,
330
+ "learning_rate": 0.0001,
331
+ "loss": 4.3528,
332
+ "step": 5400
333
+ },
334
+ {
335
+ "epoch": 4.3,
336
+ "learning_rate": 0.0001,
337
+ "loss": 4.3272,
338
+ "step": 5500
339
+ },
340
+ {
341
+ "epoch": 4.38,
342
+ "learning_rate": 0.0001,
343
+ "loss": 4.3679,
344
+ "step": 5600
345
+ },
346
+ {
347
+ "epoch": 4.45,
348
+ "learning_rate": 0.0001,
349
+ "loss": 4.3608,
350
+ "step": 5700
351
+ },
352
+ {
353
+ "epoch": 4.53,
354
+ "learning_rate": 0.0001,
355
+ "loss": 4.3262,
356
+ "step": 5800
357
+ },
358
+ {
359
+ "epoch": 4.61,
360
+ "learning_rate": 0.0001,
361
+ "loss": 4.3097,
362
+ "step": 5900
363
+ },
364
+ {
365
+ "epoch": 4.69,
366
+ "learning_rate": 0.0001,
367
+ "loss": 4.2973,
368
+ "step": 6000
369
+ },
370
+ {
371
+ "epoch": 4.77,
372
+ "learning_rate": 0.0001,
373
+ "loss": 4.2756,
374
+ "step": 6100
375
+ },
376
+ {
377
+ "epoch": 4.84,
378
+ "learning_rate": 0.0001,
379
+ "loss": 4.3174,
380
+ "step": 6200
381
+ },
382
+ {
383
+ "epoch": 4.92,
384
+ "learning_rate": 0.0001,
385
+ "loss": 4.2786,
386
+ "step": 6300
387
+ },
388
+ {
389
+ "epoch": 5.0,
390
+ "learning_rate": 0.0001,
391
+ "loss": 4.2484,
392
+ "step": 6400
393
+ },
394
+ {
395
+ "epoch": 5.08,
396
+ "learning_rate": 0.0001,
397
+ "loss": 4.2634,
398
+ "step": 6500
399
+ },
400
+ {
401
+ "epoch": 5.16,
402
+ "learning_rate": 0.0001,
403
+ "loss": 4.228,
404
+ "step": 6600
405
+ },
406
+ {
407
+ "epoch": 5.23,
408
+ "learning_rate": 0.0001,
409
+ "loss": 4.2745,
410
+ "step": 6700
411
+ },
412
+ {
413
+ "epoch": 5.31,
414
+ "learning_rate": 0.0001,
415
+ "loss": 4.2297,
416
+ "step": 6800
417
+ },
418
+ {
419
+ "epoch": 5.39,
420
+ "learning_rate": 0.0001,
421
+ "loss": 4.1989,
422
+ "step": 6900
423
+ },
424
+ {
425
+ "epoch": 5.47,
426
+ "learning_rate": 0.0001,
427
+ "loss": 4.2074,
428
+ "step": 7000
429
+ },
430
+ {
431
+ "epoch": 5.55,
432
+ "learning_rate": 0.0001,
433
+ "loss": 4.201,
434
+ "step": 7100
435
+ },
436
+ {
437
+ "epoch": 5.62,
438
+ "learning_rate": 0.0001,
439
+ "loss": 4.2109,
440
+ "step": 7200
441
+ },
442
+ {
443
+ "epoch": 5.7,
444
+ "learning_rate": 0.0001,
445
+ "loss": 4.1945,
446
+ "step": 7300
447
+ },
448
+ {
449
+ "epoch": 5.78,
450
+ "learning_rate": 0.0001,
451
+ "loss": 4.1744,
452
+ "step": 7400
453
+ },
454
+ {
455
+ "epoch": 5.86,
456
+ "learning_rate": 0.0001,
457
+ "loss": 4.1741,
458
+ "step": 7500
459
+ },
460
+ {
461
+ "epoch": 5.94,
462
+ "learning_rate": 0.0001,
463
+ "loss": 4.19,
464
+ "step": 7600
465
+ },
466
+ {
467
+ "epoch": 6.02,
468
+ "learning_rate": 0.0001,
469
+ "loss": 4.1612,
470
+ "step": 7700
471
+ },
472
+ {
473
+ "epoch": 6.09,
474
+ "learning_rate": 0.0001,
475
+ "loss": 4.1559,
476
+ "step": 7800
477
+ },
478
+ {
479
+ "epoch": 6.17,
480
+ "learning_rate": 0.0001,
481
+ "loss": 4.1378,
482
+ "step": 7900
483
+ },
484
+ {
485
+ "epoch": 6.25,
486
+ "learning_rate": 0.0001,
487
+ "loss": 4.144,
488
+ "step": 8000
489
+ },
490
+ {
491
+ "epoch": 6.33,
492
+ "learning_rate": 0.0001,
493
+ "loss": 4.1601,
494
+ "step": 8100
495
+ },
496
+ {
497
+ "epoch": 6.41,
498
+ "learning_rate": 0.0001,
499
+ "loss": 4.1339,
500
+ "step": 8200
501
+ },
502
+ {
503
+ "epoch": 6.48,
504
+ "learning_rate": 0.0001,
505
+ "loss": 4.13,
506
+ "step": 8300
507
+ },
508
+ {
509
+ "epoch": 6.56,
510
+ "learning_rate": 0.0001,
511
+ "loss": 4.0877,
512
+ "step": 8400
513
+ },
514
+ {
515
+ "epoch": 6.64,
516
+ "learning_rate": 0.0001,
517
+ "loss": 4.1022,
518
+ "step": 8500
519
+ },
520
+ {
521
+ "epoch": 6.72,
522
+ "learning_rate": 0.0001,
523
+ "loss": 4.1034,
524
+ "step": 8600
525
+ },
526
+ {
527
+ "epoch": 6.8,
528
+ "learning_rate": 0.0001,
529
+ "loss": 4.091,
530
+ "step": 8700
531
+ },
532
+ {
533
+ "epoch": 6.88,
534
+ "learning_rate": 0.0001,
535
+ "loss": 4.0914,
536
+ "step": 8800
537
+ },
538
+ {
539
+ "epoch": 6.95,
540
+ "learning_rate": 0.0001,
541
+ "loss": 4.0937,
542
+ "step": 8900
543
+ },
544
+ {
545
+ "epoch": 7.03,
546
+ "learning_rate": 0.0001,
547
+ "loss": 4.0897,
548
+ "step": 9000
549
+ },
550
+ {
551
+ "epoch": 7.11,
552
+ "learning_rate": 0.0001,
553
+ "loss": 4.0844,
554
+ "step": 9100
555
+ },
556
+ {
557
+ "epoch": 7.19,
558
+ "learning_rate": 0.0001,
559
+ "loss": 4.0574,
560
+ "step": 9200
561
+ },
562
+ {
563
+ "epoch": 7.27,
564
+ "learning_rate": 0.0001,
565
+ "loss": 4.0837,
566
+ "step": 9300
567
+ },
568
+ {
569
+ "epoch": 7.34,
570
+ "learning_rate": 0.0001,
571
+ "loss": 4.0766,
572
+ "step": 9400
573
+ },
574
+ {
575
+ "epoch": 7.42,
576
+ "learning_rate": 0.0001,
577
+ "loss": 4.066,
578
+ "step": 9500
579
+ },
580
+ {
581
+ "epoch": 7.5,
582
+ "learning_rate": 0.0001,
583
+ "loss": 3.9944,
584
+ "step": 9600
585
+ },
586
+ {
587
+ "epoch": 7.58,
588
+ "learning_rate": 0.0001,
589
+ "loss": 4.0126,
590
+ "step": 9700
591
+ },
592
+ {
593
+ "epoch": 7.66,
594
+ "learning_rate": 0.0001,
595
+ "loss": 4.0273,
596
+ "step": 9800
597
+ },
598
+ {
599
+ "epoch": 7.73,
600
+ "learning_rate": 0.0001,
601
+ "loss": 4.0706,
602
+ "step": 9900
603
+ },
604
+ {
605
+ "epoch": 7.81,
606
+ "learning_rate": 0.0001,
607
+ "loss": 4.0008,
608
+ "step": 10000
609
+ },
610
+ {
611
+ "epoch": 7.89,
612
+ "learning_rate": 0.0001,
613
+ "loss": 4.0323,
614
+ "step": 10100
615
+ },
616
+ {
617
+ "epoch": 7.97,
618
+ "learning_rate": 0.0001,
619
+ "loss": 3.9864,
620
+ "step": 10200
621
+ },
622
+ {
623
+ "epoch": 8.05,
624
+ "learning_rate": 0.0001,
625
+ "loss": 4.0265,
626
+ "step": 10300
627
+ },
628
+ {
629
+ "epoch": 8.12,
630
+ "learning_rate": 0.0001,
631
+ "loss": 3.9874,
632
+ "step": 10400
633
+ },
634
+ {
635
+ "epoch": 8.2,
636
+ "learning_rate": 0.0001,
637
+ "loss": 3.9839,
638
+ "step": 10500
639
+ },
640
+ {
641
+ "epoch": 8.28,
642
+ "learning_rate": 0.0001,
643
+ "loss": 3.9583,
644
+ "step": 10600
645
+ },
646
+ {
647
+ "epoch": 8.36,
648
+ "learning_rate": 0.0001,
649
+ "loss": 4.007,
650
+ "step": 10700
651
+ },
652
+ {
653
+ "epoch": 8.44,
654
+ "learning_rate": 0.0001,
655
+ "loss": 3.9567,
656
+ "step": 10800
657
+ },
658
+ {
659
+ "epoch": 8.52,
660
+ "learning_rate": 0.0001,
661
+ "loss": 3.9888,
662
+ "step": 10900
663
+ },
664
+ {
665
+ "epoch": 8.59,
666
+ "learning_rate": 0.0001,
667
+ "loss": 3.9993,
668
+ "step": 11000
669
+ },
670
+ {
671
+ "epoch": 8.67,
672
+ "learning_rate": 0.0001,
673
+ "loss": 3.9914,
674
+ "step": 11100
675
+ },
676
+ {
677
+ "epoch": 8.75,
678
+ "learning_rate": 0.0001,
679
+ "loss": 4.0094,
680
+ "step": 11200
681
+ },
682
+ {
683
+ "epoch": 8.83,
684
+ "learning_rate": 0.0001,
685
+ "loss": 3.9338,
686
+ "step": 11300
687
+ },
688
+ {
689
+ "epoch": 8.91,
690
+ "learning_rate": 0.0001,
691
+ "loss": 3.9472,
692
+ "step": 11400
693
+ },
694
+ {
695
+ "epoch": 8.98,
696
+ "learning_rate": 0.0001,
697
+ "loss": 3.9595,
698
+ "step": 11500
699
+ },
700
+ {
701
+ "epoch": 9.06,
702
+ "learning_rate": 0.0001,
703
+ "loss": 3.9545,
704
+ "step": 11600
705
+ },
706
+ {
707
+ "epoch": 9.14,
708
+ "learning_rate": 0.0001,
709
+ "loss": 3.993,
710
+ "step": 11700
711
+ },
712
+ {
713
+ "epoch": 9.22,
714
+ "learning_rate": 0.0001,
715
+ "loss": 3.895,
716
+ "step": 11800
717
+ },
718
+ {
719
+ "epoch": 9.3,
720
+ "learning_rate": 0.0001,
721
+ "loss": 3.9527,
722
+ "step": 11900
723
+ },
724
+ {
725
+ "epoch": 9.38,
726
+ "learning_rate": 0.0001,
727
+ "loss": 3.9205,
728
+ "step": 12000
729
+ },
730
+ {
731
+ "epoch": 9.45,
732
+ "learning_rate": 0.0001,
733
+ "loss": 3.914,
734
+ "step": 12100
735
+ },
736
+ {
737
+ "epoch": 9.53,
738
+ "learning_rate": 0.0001,
739
+ "loss": 3.8872,
740
+ "step": 12200
741
+ },
742
+ {
743
+ "epoch": 9.61,
744
+ "learning_rate": 0.0001,
745
+ "loss": 3.9171,
746
+ "step": 12300
747
+ },
748
+ {
749
+ "epoch": 9.69,
750
+ "learning_rate": 0.0001,
751
+ "loss": 3.9148,
752
+ "step": 12400
753
+ }
754
+ ],
755
+ "max_steps": 12800,
756
+ "num_train_epochs": 10,
757
+ "total_flos": 9292082437785600.0,
758
+ "trial_name": null,
759
+ "trial_params": null
760
+ }
checkpoint-12420/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e683e2614c4dcf1ee5d4bfe9622ed2cfdbf3faf9319855e4a9b393368f21e44
3
+ size 3899
checkpoint-12450/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: fp4
15
+ - bnb_4bit_use_double_quant: False
16
+ - bnb_4bit_compute_dtype: float32
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.5.0.dev0
checkpoint-12450/adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "models/TinyStories-33M",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 8,
11
+ "lora_dropout": 0.0,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 16,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
checkpoint-12450/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:31656b0458422251ad42a99d942ee864bd478e658dc330893b815d1bb5af7fb4
3
+ size 792353
checkpoint-12450/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ad54a640ed5e53747f52caa10179f60dd135cc0ae48257cdf042352a7a6e6ee0
3
+ size 1586873
checkpoint-12450/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8bb3860d98363e44bc4704fb9458deec927fd96997d1d09c80639070db959d22
3
+ size 14575
checkpoint-12450/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a3d27439df90a9a7352314b9b8e368610f9dd675c06874da737b447cd1bc3435
3
+ size 627
checkpoint-12450/trainer_state.json ADDED
@@ -0,0 +1,760 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 9.7265625,
5
+ "global_step": 12450,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.08,
12
+ "learning_rate": 0.0001,
13
+ "loss": 10.386,
14
+ "step": 100
15
+ },
16
+ {
17
+ "epoch": 0.16,
18
+ "learning_rate": 0.0001,
19
+ "loss": 7.4061,
20
+ "step": 200
21
+ },
22
+ {
23
+ "epoch": 0.23,
24
+ "learning_rate": 0.0001,
25
+ "loss": 6.4112,
26
+ "step": 300
27
+ },
28
+ {
29
+ "epoch": 0.31,
30
+ "learning_rate": 0.0001,
31
+ "loss": 6.091,
32
+ "step": 400
33
+ },
34
+ {
35
+ "epoch": 0.39,
36
+ "learning_rate": 0.0001,
37
+ "loss": 5.8571,
38
+ "step": 500
39
+ },
40
+ {
41
+ "epoch": 0.47,
42
+ "learning_rate": 0.0001,
43
+ "loss": 5.7144,
44
+ "step": 600
45
+ },
46
+ {
47
+ "epoch": 0.55,
48
+ "learning_rate": 0.0001,
49
+ "loss": 5.6228,
50
+ "step": 700
51
+ },
52
+ {
53
+ "epoch": 0.62,
54
+ "learning_rate": 0.0001,
55
+ "loss": 5.546,
56
+ "step": 800
57
+ },
58
+ {
59
+ "epoch": 0.7,
60
+ "learning_rate": 0.0001,
61
+ "loss": 5.4904,
62
+ "step": 900
63
+ },
64
+ {
65
+ "epoch": 0.78,
66
+ "learning_rate": 0.0001,
67
+ "loss": 5.4413,
68
+ "step": 1000
69
+ },
70
+ {
71
+ "epoch": 0.86,
72
+ "learning_rate": 0.0001,
73
+ "loss": 5.3603,
74
+ "step": 1100
75
+ },
76
+ {
77
+ "epoch": 0.94,
78
+ "learning_rate": 0.0001,
79
+ "loss": 5.3346,
80
+ "step": 1200
81
+ },
82
+ {
83
+ "epoch": 1.02,
84
+ "learning_rate": 0.0001,
85
+ "loss": 5.2744,
86
+ "step": 1300
87
+ },
88
+ {
89
+ "epoch": 1.09,
90
+ "learning_rate": 0.0001,
91
+ "loss": 5.231,
92
+ "step": 1400
93
+ },
94
+ {
95
+ "epoch": 1.17,
96
+ "learning_rate": 0.0001,
97
+ "loss": 5.1782,
98
+ "step": 1500
99
+ },
100
+ {
101
+ "epoch": 1.25,
102
+ "learning_rate": 0.0001,
103
+ "loss": 5.1613,
104
+ "step": 1600
105
+ },
106
+ {
107
+ "epoch": 1.33,
108
+ "learning_rate": 0.0001,
109
+ "loss": 5.1091,
110
+ "step": 1700
111
+ },
112
+ {
113
+ "epoch": 1.41,
114
+ "learning_rate": 0.0001,
115
+ "loss": 5.0939,
116
+ "step": 1800
117
+ },
118
+ {
119
+ "epoch": 1.48,
120
+ "learning_rate": 0.0001,
121
+ "loss": 5.0667,
122
+ "step": 1900
123
+ },
124
+ {
125
+ "epoch": 1.56,
126
+ "learning_rate": 0.0001,
127
+ "loss": 5.0054,
128
+ "step": 2000
129
+ },
130
+ {
131
+ "epoch": 1.64,
132
+ "learning_rate": 0.0001,
133
+ "loss": 4.96,
134
+ "step": 2100
135
+ },
136
+ {
137
+ "epoch": 1.72,
138
+ "learning_rate": 0.0001,
139
+ "loss": 4.9301,
140
+ "step": 2200
141
+ },
142
+ {
143
+ "epoch": 1.8,
144
+ "learning_rate": 0.0001,
145
+ "loss": 4.9302,
146
+ "step": 2300
147
+ },
148
+ {
149
+ "epoch": 1.88,
150
+ "learning_rate": 0.0001,
151
+ "loss": 4.9129,
152
+ "step": 2400
153
+ },
154
+ {
155
+ "epoch": 1.95,
156
+ "learning_rate": 0.0001,
157
+ "loss": 4.8627,
158
+ "step": 2500
159
+ },
160
+ {
161
+ "epoch": 2.03,
162
+ "learning_rate": 0.0001,
163
+ "loss": 4.8336,
164
+ "step": 2600
165
+ },
166
+ {
167
+ "epoch": 2.11,
168
+ "learning_rate": 0.0001,
169
+ "loss": 4.8018,
170
+ "step": 2700
171
+ },
172
+ {
173
+ "epoch": 2.19,
174
+ "learning_rate": 0.0001,
175
+ "loss": 4.7479,
176
+ "step": 2800
177
+ },
178
+ {
179
+ "epoch": 2.27,
180
+ "learning_rate": 0.0001,
181
+ "loss": 4.7572,
182
+ "step": 2900
183
+ },
184
+ {
185
+ "epoch": 2.34,
186
+ "learning_rate": 0.0001,
187
+ "loss": 4.759,
188
+ "step": 3000
189
+ },
190
+ {
191
+ "epoch": 2.42,
192
+ "learning_rate": 0.0001,
193
+ "loss": 4.7441,
194
+ "step": 3100
195
+ },
196
+ {
197
+ "epoch": 2.5,
198
+ "learning_rate": 0.0001,
199
+ "loss": 4.7195,
200
+ "step": 3200
201
+ },
202
+ {
203
+ "epoch": 2.58,
204
+ "learning_rate": 0.0001,
205
+ "loss": 4.6821,
206
+ "step": 3300
207
+ },
208
+ {
209
+ "epoch": 2.66,
210
+ "learning_rate": 0.0001,
211
+ "loss": 4.6587,
212
+ "step": 3400
213
+ },
214
+ {
215
+ "epoch": 2.73,
216
+ "learning_rate": 0.0001,
217
+ "loss": 4.6228,
218
+ "step": 3500
219
+ },
220
+ {
221
+ "epoch": 2.81,
222
+ "learning_rate": 0.0001,
223
+ "loss": 4.6248,
224
+ "step": 3600
225
+ },
226
+ {
227
+ "epoch": 2.89,
228
+ "learning_rate": 0.0001,
229
+ "loss": 4.5801,
230
+ "step": 3700
231
+ },
232
+ {
233
+ "epoch": 2.97,
234
+ "learning_rate": 0.0001,
235
+ "loss": 4.5934,
236
+ "step": 3800
237
+ },
238
+ {
239
+ "epoch": 3.05,
240
+ "learning_rate": 0.0001,
241
+ "loss": 4.5781,
242
+ "step": 3900
243
+ },
244
+ {
245
+ "epoch": 3.12,
246
+ "learning_rate": 0.0001,
247
+ "loss": 4.5763,
248
+ "step": 4000
249
+ },
250
+ {
251
+ "epoch": 3.2,
252
+ "learning_rate": 0.0001,
253
+ "loss": 4.5598,
254
+ "step": 4100
255
+ },
256
+ {
257
+ "epoch": 3.28,
258
+ "learning_rate": 0.0001,
259
+ "loss": 4.4796,
260
+ "step": 4200
261
+ },
262
+ {
263
+ "epoch": 3.36,
264
+ "learning_rate": 0.0001,
265
+ "loss": 4.5204,
266
+ "step": 4300
267
+ },
268
+ {
269
+ "epoch": 3.44,
270
+ "learning_rate": 0.0001,
271
+ "loss": 4.4858,
272
+ "step": 4400
273
+ },
274
+ {
275
+ "epoch": 3.52,
276
+ "learning_rate": 0.0001,
277
+ "loss": 4.4882,
278
+ "step": 4500
279
+ },
280
+ {
281
+ "epoch": 3.59,
282
+ "learning_rate": 0.0001,
283
+ "loss": 4.4852,
284
+ "step": 4600
285
+ },
286
+ {
287
+ "epoch": 3.67,
288
+ "learning_rate": 0.0001,
289
+ "loss": 4.4302,
290
+ "step": 4700
291
+ },
292
+ {
293
+ "epoch": 3.75,
294
+ "learning_rate": 0.0001,
295
+ "loss": 4.4267,
296
+ "step": 4800
297
+ },
298
+ {
299
+ "epoch": 3.83,
300
+ "learning_rate": 0.0001,
301
+ "loss": 4.4298,
302
+ "step": 4900
303
+ },
304
+ {
305
+ "epoch": 3.91,
306
+ "learning_rate": 0.0001,
307
+ "loss": 4.3892,
308
+ "step": 5000
309
+ },
310
+ {
311
+ "epoch": 3.98,
312
+ "learning_rate": 0.0001,
313
+ "loss": 4.3886,
314
+ "step": 5100
315
+ },
316
+ {
317
+ "epoch": 4.06,
318
+ "learning_rate": 0.0001,
319
+ "loss": 4.3915,
320
+ "step": 5200
321
+ },
322
+ {
323
+ "epoch": 4.14,
324
+ "learning_rate": 0.0001,
325
+ "loss": 4.3706,
326
+ "step": 5300
327
+ },
328
+ {
329
+ "epoch": 4.22,
330
+ "learning_rate": 0.0001,
331
+ "loss": 4.3528,
332
+ "step": 5400
333
+ },
334
+ {
335
+ "epoch": 4.3,
336
+ "learning_rate": 0.0001,
337
+ "loss": 4.3272,
338
+ "step": 5500
339
+ },
340
+ {
341
+ "epoch": 4.38,
342
+ "learning_rate": 0.0001,
343
+ "loss": 4.3679,
344
+ "step": 5600
345
+ },
346
+ {
347
+ "epoch": 4.45,
348
+ "learning_rate": 0.0001,
349
+ "loss": 4.3608,
350
+ "step": 5700
351
+ },
352
+ {
353
+ "epoch": 4.53,
354
+ "learning_rate": 0.0001,
355
+ "loss": 4.3262,
356
+ "step": 5800
357
+ },
358
+ {
359
+ "epoch": 4.61,
360
+ "learning_rate": 0.0001,
361
+ "loss": 4.3097,
362
+ "step": 5900
363
+ },
364
+ {
365
+ "epoch": 4.69,
366
+ "learning_rate": 0.0001,
367
+ "loss": 4.2973,
368
+ "step": 6000
369
+ },
370
+ {
371
+ "epoch": 4.77,
372
+ "learning_rate": 0.0001,
373
+ "loss": 4.2756,
374
+ "step": 6100
375
+ },
376
+ {
377
+ "epoch": 4.84,
378
+ "learning_rate": 0.0001,
379
+ "loss": 4.3174,
380
+ "step": 6200
381
+ },
382
+ {
383
+ "epoch": 4.92,
384
+ "learning_rate": 0.0001,
385
+ "loss": 4.2786,
386
+ "step": 6300
387
+ },
388
+ {
389
+ "epoch": 5.0,
390
+ "learning_rate": 0.0001,
391
+ "loss": 4.2484,
392
+ "step": 6400
393
+ },
394
+ {
395
+ "epoch": 5.08,
396
+ "learning_rate": 0.0001,
397
+ "loss": 4.2634,
398
+ "step": 6500
399
+ },
400
+ {
401
+ "epoch": 5.16,
402
+ "learning_rate": 0.0001,
403
+ "loss": 4.228,
404
+ "step": 6600
405
+ },
406
+ {
407
+ "epoch": 5.23,
408
+ "learning_rate": 0.0001,
409
+ "loss": 4.2745,
410
+ "step": 6700
411
+ },
412
+ {
413
+ "epoch": 5.31,
414
+ "learning_rate": 0.0001,
415
+ "loss": 4.2297,
416
+ "step": 6800
417
+ },
418
+ {
419
+ "epoch": 5.39,
420
+ "learning_rate": 0.0001,
421
+ "loss": 4.1989,
422
+ "step": 6900
423
+ },
424
+ {
425
+ "epoch": 5.47,
426
+ "learning_rate": 0.0001,
427
+ "loss": 4.2074,
428
+ "step": 7000
429
+ },
430
+ {
431
+ "epoch": 5.55,
432
+ "learning_rate": 0.0001,
433
+ "loss": 4.201,
434
+ "step": 7100
435
+ },
436
+ {
437
+ "epoch": 5.62,
438
+ "learning_rate": 0.0001,
439
+ "loss": 4.2109,
440
+ "step": 7200
441
+ },
442
+ {
443
+ "epoch": 5.7,
444
+ "learning_rate": 0.0001,
445
+ "loss": 4.1945,
446
+ "step": 7300
447
+ },
448
+ {
449
+ "epoch": 5.78,
450
+ "learning_rate": 0.0001,
451
+ "loss": 4.1744,
452
+ "step": 7400
453
+ },
454
+ {
455
+ "epoch": 5.86,
456
+ "learning_rate": 0.0001,
457
+ "loss": 4.1741,
458
+ "step": 7500
459
+ },
460
+ {
461
+ "epoch": 5.94,
462
+ "learning_rate": 0.0001,
463
+ "loss": 4.19,
464
+ "step": 7600
465
+ },
466
+ {
467
+ "epoch": 6.02,
468
+ "learning_rate": 0.0001,
469
+ "loss": 4.1612,
470
+ "step": 7700
471
+ },
472
+ {
473
+ "epoch": 6.09,
474
+ "learning_rate": 0.0001,
475
+ "loss": 4.1559,
476
+ "step": 7800
477
+ },
478
+ {
479
+ "epoch": 6.17,
480
+ "learning_rate": 0.0001,
481
+ "loss": 4.1378,
482
+ "step": 7900
483
+ },
484
+ {
485
+ "epoch": 6.25,
486
+ "learning_rate": 0.0001,
487
+ "loss": 4.144,
488
+ "step": 8000
489
+ },
490
+ {
491
+ "epoch": 6.33,
492
+ "learning_rate": 0.0001,
493
+ "loss": 4.1601,
494
+ "step": 8100
495
+ },
496
+ {
497
+ "epoch": 6.41,
498
+ "learning_rate": 0.0001,
499
+ "loss": 4.1339,
500
+ "step": 8200
501
+ },
502
+ {
503
+ "epoch": 6.48,
504
+ "learning_rate": 0.0001,
505
+ "loss": 4.13,
506
+ "step": 8300
507
+ },
508
+ {
509
+ "epoch": 6.56,
510
+ "learning_rate": 0.0001,
511
+ "loss": 4.0877,
512
+ "step": 8400
513
+ },
514
+ {
515
+ "epoch": 6.64,
516
+ "learning_rate": 0.0001,
517
+ "loss": 4.1022,
518
+ "step": 8500
519
+ },
520
+ {
521
+ "epoch": 6.72,
522
+ "learning_rate": 0.0001,
523
+ "loss": 4.1034,
524
+ "step": 8600
525
+ },
526
+ {
527
+ "epoch": 6.8,
528
+ "learning_rate": 0.0001,
529
+ "loss": 4.091,
530
+ "step": 8700
531
+ },
532
+ {
533
+ "epoch": 6.88,
534
+ "learning_rate": 0.0001,
535
+ "loss": 4.0914,
536
+ "step": 8800
537
+ },
538
+ {
539
+ "epoch": 6.95,
540
+ "learning_rate": 0.0001,
541
+ "loss": 4.0937,
542
+ "step": 8900
543
+ },
544
+ {
545
+ "epoch": 7.03,
546
+ "learning_rate": 0.0001,
547
+ "loss": 4.0897,
548
+ "step": 9000
549
+ },
550
+ {
551
+ "epoch": 7.11,
552
+ "learning_rate": 0.0001,
553
+ "loss": 4.0844,
554
+ "step": 9100
555
+ },
556
+ {
557
+ "epoch": 7.19,
558
+ "learning_rate": 0.0001,
559
+ "loss": 4.0574,
560
+ "step": 9200
561
+ },
562
+ {
563
+ "epoch": 7.27,
564
+ "learning_rate": 0.0001,
565
+ "loss": 4.0837,
566
+ "step": 9300
567
+ },
568
+ {
569
+ "epoch": 7.34,
570
+ "learning_rate": 0.0001,
571
+ "loss": 4.0766,
572
+ "step": 9400
573
+ },
574
+ {
575
+ "epoch": 7.42,
576
+ "learning_rate": 0.0001,
577
+ "loss": 4.066,
578
+ "step": 9500
579
+ },
580
+ {
581
+ "epoch": 7.5,
582
+ "learning_rate": 0.0001,
583
+ "loss": 3.9944,
584
+ "step": 9600
585
+ },
586
+ {
587
+ "epoch": 7.58,
588
+ "learning_rate": 0.0001,
589
+ "loss": 4.0126,
590
+ "step": 9700
591
+ },
592
+ {
593
+ "epoch": 7.66,
594
+ "learning_rate": 0.0001,
595
+ "loss": 4.0273,
596
+ "step": 9800
597
+ },
598
+ {
599
+ "epoch": 7.73,
600
+ "learning_rate": 0.0001,
601
+ "loss": 4.0706,
602
+ "step": 9900
603
+ },
604
+ {
605
+ "epoch": 7.81,
606
+ "learning_rate": 0.0001,
607
+ "loss": 4.0008,
608
+ "step": 10000
609
+ },
610
+ {
611
+ "epoch": 7.89,
612
+ "learning_rate": 0.0001,
613
+ "loss": 4.0323,
614
+ "step": 10100
615
+ },
616
+ {
617
+ "epoch": 7.97,
618
+ "learning_rate": 0.0001,
619
+ "loss": 3.9864,
620
+ "step": 10200
621
+ },
622
+ {
623
+ "epoch": 8.05,
624
+ "learning_rate": 0.0001,
625
+ "loss": 4.0265,
626
+ "step": 10300
627
+ },
628
+ {
629
+ "epoch": 8.12,
630
+ "learning_rate": 0.0001,
631
+ "loss": 3.9874,
632
+ "step": 10400
633
+ },
634
+ {
635
+ "epoch": 8.2,
636
+ "learning_rate": 0.0001,
637
+ "loss": 3.9839,
638
+ "step": 10500
639
+ },
640
+ {
641
+ "epoch": 8.28,
642
+ "learning_rate": 0.0001,
643
+ "loss": 3.9583,
644
+ "step": 10600
645
+ },
646
+ {
647
+ "epoch": 8.36,
648
+ "learning_rate": 0.0001,
649
+ "loss": 4.007,
650
+ "step": 10700
651
+ },
652
+ {
653
+ "epoch": 8.44,
654
+ "learning_rate": 0.0001,
655
+ "loss": 3.9567,
656
+ "step": 10800
657
+ },
658
+ {
659
+ "epoch": 8.52,
660
+ "learning_rate": 0.0001,
661
+ "loss": 3.9888,
662
+ "step": 10900
663
+ },
664
+ {
665
+ "epoch": 8.59,
666
+ "learning_rate": 0.0001,
667
+ "loss": 3.9993,
668
+ "step": 11000
669
+ },
670
+ {
671
+ "epoch": 8.67,
672
+ "learning_rate": 0.0001,
673
+ "loss": 3.9914,
674
+ "step": 11100
675
+ },
676
+ {
677
+ "epoch": 8.75,
678
+ "learning_rate": 0.0001,
679
+ "loss": 4.0094,
680
+ "step": 11200
681
+ },
682
+ {
683
+ "epoch": 8.83,
684
+ "learning_rate": 0.0001,
685
+ "loss": 3.9338,
686
+ "step": 11300
687
+ },
688
+ {
689
+ "epoch": 8.91,
690
+ "learning_rate": 0.0001,
691
+ "loss": 3.9472,
692
+ "step": 11400
693
+ },
694
+ {
695
+ "epoch": 8.98,
696
+ "learning_rate": 0.0001,
697
+ "loss": 3.9595,
698
+ "step": 11500
699
+ },
700
+ {
701
+ "epoch": 9.06,
702
+ "learning_rate": 0.0001,
703
+ "loss": 3.9545,
704
+ "step": 11600
705
+ },
706
+ {
707
+ "epoch": 9.14,
708
+ "learning_rate": 0.0001,
709
+ "loss": 3.993,
710
+ "step": 11700
711
+ },
712
+ {
713
+ "epoch": 9.22,
714
+ "learning_rate": 0.0001,
715
+ "loss": 3.895,
716
+ "step": 11800
717
+ },
718
+ {
719
+ "epoch": 9.3,
720
+ "learning_rate": 0.0001,
721
+ "loss": 3.9527,
722
+ "step": 11900
723
+ },
724
+ {
725
+ "epoch": 9.38,
726
+ "learning_rate": 0.0001,
727
+ "loss": 3.9205,
728
+ "step": 12000
729
+ },
730
+ {
731
+ "epoch": 9.45,
732
+ "learning_rate": 0.0001,
733
+ "loss": 3.914,
734
+ "step": 12100
735
+ },
736
+ {
737
+ "epoch": 9.53,
738
+ "learning_rate": 0.0001,
739
+ "loss": 3.8872,
740
+ "step": 12200
741
+ },
742
+ {
743
+ "epoch": 9.61,
744
+ "learning_rate": 0.0001,
745
+ "loss": 3.9171,
746
+ "step": 12300
747
+ },
748
+ {
749
+ "epoch": 9.69,
750
+ "learning_rate": 0.0001,
751
+ "loss": 3.9148,
752
+ "step": 12400
753
+ }
754
+ ],
755
+ "max_steps": 12800,
756
+ "num_train_epochs": 10,
757
+ "total_flos": 9314514190632960.0,
758
+ "trial_name": null,
759
+ "trial_params": null
760
+ }
checkpoint-12450/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e683e2614c4dcf1ee5d4bfe9622ed2cfdbf3faf9319855e4a9b393368f21e44
3
+ size 3899
checkpoint-12480/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: fp4
15
+ - bnb_4bit_use_double_quant: False
16
+ - bnb_4bit_compute_dtype: float32
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.5.0.dev0
checkpoint-12480/adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "models/TinyStories-33M",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 8,
11
+ "lora_dropout": 0.0,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 16,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
checkpoint-12480/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e59da02aba0ecff94c22170bd5085bbb8e5bac87f5689bde7d13318ccaa683d4
3
+ size 792353
checkpoint-12480/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:78af0f47d42de1be9f4882eb373ce9f1a2130d3396b4ad2d8ee94886ef73ad3a
3
+ size 1586873
checkpoint-12480/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8bb3860d98363e44bc4704fb9458deec927fd96997d1d09c80639070db959d22
3
+ size 14575
checkpoint-12480/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f07aed61c9e1af3e6872c4649dbcbdb2c02eaa0deb0ce84f1b9db4bec9454439
3
+ size 627
checkpoint-12480/trainer_state.json ADDED
@@ -0,0 +1,760 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 9.75,
5
+ "global_step": 12480,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.08,
12
+ "learning_rate": 0.0001,
13
+ "loss": 10.386,
14
+ "step": 100
15
+ },
16
+ {
17
+ "epoch": 0.16,
18
+ "learning_rate": 0.0001,
19
+ "loss": 7.4061,
20
+ "step": 200
21
+ },
22
+ {
23
+ "epoch": 0.23,
24
+ "learning_rate": 0.0001,
25
+ "loss": 6.4112,
26
+ "step": 300
27
+ },
28
+ {
29
+ "epoch": 0.31,
30
+ "learning_rate": 0.0001,
31
+ "loss": 6.091,
32
+ "step": 400
33
+ },
34
+ {
35
+ "epoch": 0.39,
36
+ "learning_rate": 0.0001,
37
+ "loss": 5.8571,
38
+ "step": 500
39
+ },
40
+ {
41
+ "epoch": 0.47,
42
+ "learning_rate": 0.0001,
43
+ "loss": 5.7144,
44
+ "step": 600
45
+ },
46
+ {
47
+ "epoch": 0.55,
48
+ "learning_rate": 0.0001,
49
+ "loss": 5.6228,
50
+ "step": 700
51
+ },
52
+ {
53
+ "epoch": 0.62,
54
+ "learning_rate": 0.0001,
55
+ "loss": 5.546,
56
+ "step": 800
57
+ },
58
+ {
59
+ "epoch": 0.7,
60
+ "learning_rate": 0.0001,
61
+ "loss": 5.4904,
62
+ "step": 900
63
+ },
64
+ {
65
+ "epoch": 0.78,
66
+ "learning_rate": 0.0001,
67
+ "loss": 5.4413,
68
+ "step": 1000
69
+ },
70
+ {
71
+ "epoch": 0.86,
72
+ "learning_rate": 0.0001,
73
+ "loss": 5.3603,
74
+ "step": 1100
75
+ },
76
+ {
77
+ "epoch": 0.94,
78
+ "learning_rate": 0.0001,
79
+ "loss": 5.3346,
80
+ "step": 1200
81
+ },
82
+ {
83
+ "epoch": 1.02,
84
+ "learning_rate": 0.0001,
85
+ "loss": 5.2744,
86
+ "step": 1300
87
+ },
88
+ {
89
+ "epoch": 1.09,
90
+ "learning_rate": 0.0001,
91
+ "loss": 5.231,
92
+ "step": 1400
93
+ },
94
+ {
95
+ "epoch": 1.17,
96
+ "learning_rate": 0.0001,
97
+ "loss": 5.1782,
98
+ "step": 1500
99
+ },
100
+ {
101
+ "epoch": 1.25,
102
+ "learning_rate": 0.0001,
103
+ "loss": 5.1613,
104
+ "step": 1600
105
+ },
106
+ {
107
+ "epoch": 1.33,
108
+ "learning_rate": 0.0001,
109
+ "loss": 5.1091,
110
+ "step": 1700
111
+ },
112
+ {
113
+ "epoch": 1.41,
114
+ "learning_rate": 0.0001,
115
+ "loss": 5.0939,
116
+ "step": 1800
117
+ },
118
+ {
119
+ "epoch": 1.48,
120
+ "learning_rate": 0.0001,
121
+ "loss": 5.0667,
122
+ "step": 1900
123
+ },
124
+ {
125
+ "epoch": 1.56,
126
+ "learning_rate": 0.0001,
127
+ "loss": 5.0054,
128
+ "step": 2000
129
+ },
130
+ {
131
+ "epoch": 1.64,
132
+ "learning_rate": 0.0001,
133
+ "loss": 4.96,
134
+ "step": 2100
135
+ },
136
+ {
137
+ "epoch": 1.72,
138
+ "learning_rate": 0.0001,
139
+ "loss": 4.9301,
140
+ "step": 2200
141
+ },
142
+ {
143
+ "epoch": 1.8,
144
+ "learning_rate": 0.0001,
145
+ "loss": 4.9302,
146
+ "step": 2300
147
+ },
148
+ {
149
+ "epoch": 1.88,
150
+ "learning_rate": 0.0001,
151
+ "loss": 4.9129,
152
+ "step": 2400
153
+ },
154
+ {
155
+ "epoch": 1.95,
156
+ "learning_rate": 0.0001,
157
+ "loss": 4.8627,
158
+ "step": 2500
159
+ },
160
+ {
161
+ "epoch": 2.03,
162
+ "learning_rate": 0.0001,
163
+ "loss": 4.8336,
164
+ "step": 2600
165
+ },
166
+ {
167
+ "epoch": 2.11,
168
+ "learning_rate": 0.0001,
169
+ "loss": 4.8018,
170
+ "step": 2700
171
+ },
172
+ {
173
+ "epoch": 2.19,
174
+ "learning_rate": 0.0001,
175
+ "loss": 4.7479,
176
+ "step": 2800
177
+ },
178
+ {
179
+ "epoch": 2.27,
180
+ "learning_rate": 0.0001,
181
+ "loss": 4.7572,
182
+ "step": 2900
183
+ },
184
+ {
185
+ "epoch": 2.34,
186
+ "learning_rate": 0.0001,
187
+ "loss": 4.759,
188
+ "step": 3000
189
+ },
190
+ {
191
+ "epoch": 2.42,
192
+ "learning_rate": 0.0001,
193
+ "loss": 4.7441,
194
+ "step": 3100
195
+ },
196
+ {
197
+ "epoch": 2.5,
198
+ "learning_rate": 0.0001,
199
+ "loss": 4.7195,
200
+ "step": 3200
201
+ },
202
+ {
203
+ "epoch": 2.58,
204
+ "learning_rate": 0.0001,
205
+ "loss": 4.6821,
206
+ "step": 3300
207
+ },
208
+ {
209
+ "epoch": 2.66,
210
+ "learning_rate": 0.0001,
211
+ "loss": 4.6587,
212
+ "step": 3400
213
+ },
214
+ {
215
+ "epoch": 2.73,
216
+ "learning_rate": 0.0001,
217
+ "loss": 4.6228,
218
+ "step": 3500
219
+ },
220
+ {
221
+ "epoch": 2.81,
222
+ "learning_rate": 0.0001,
223
+ "loss": 4.6248,
224
+ "step": 3600
225
+ },
226
+ {
227
+ "epoch": 2.89,
228
+ "learning_rate": 0.0001,
229
+ "loss": 4.5801,
230
+ "step": 3700
231
+ },
232
+ {
233
+ "epoch": 2.97,
234
+ "learning_rate": 0.0001,
235
+ "loss": 4.5934,
236
+ "step": 3800
237
+ },
238
+ {
239
+ "epoch": 3.05,
240
+ "learning_rate": 0.0001,
241
+ "loss": 4.5781,
242
+ "step": 3900
243
+ },
244
+ {
245
+ "epoch": 3.12,
246
+ "learning_rate": 0.0001,
247
+ "loss": 4.5763,
248
+ "step": 4000
249
+ },
250
+ {
251
+ "epoch": 3.2,
252
+ "learning_rate": 0.0001,
253
+ "loss": 4.5598,
254
+ "step": 4100
255
+ },
256
+ {
257
+ "epoch": 3.28,
258
+ "learning_rate": 0.0001,
259
+ "loss": 4.4796,
260
+ "step": 4200
261
+ },
262
+ {
263
+ "epoch": 3.36,
264
+ "learning_rate": 0.0001,
265
+ "loss": 4.5204,
266
+ "step": 4300
267
+ },
268
+ {
269
+ "epoch": 3.44,
270
+ "learning_rate": 0.0001,
271
+ "loss": 4.4858,
272
+ "step": 4400
273
+ },
274
+ {
275
+ "epoch": 3.52,
276
+ "learning_rate": 0.0001,
277
+ "loss": 4.4882,
278
+ "step": 4500
279
+ },
280
+ {
281
+ "epoch": 3.59,
282
+ "learning_rate": 0.0001,
283
+ "loss": 4.4852,
284
+ "step": 4600
285
+ },
286
+ {
287
+ "epoch": 3.67,
288
+ "learning_rate": 0.0001,
289
+ "loss": 4.4302,
290
+ "step": 4700
291
+ },
292
+ {
293
+ "epoch": 3.75,
294
+ "learning_rate": 0.0001,
295
+ "loss": 4.4267,
296
+ "step": 4800
297
+ },
298
+ {
299
+ "epoch": 3.83,
300
+ "learning_rate": 0.0001,
301
+ "loss": 4.4298,
302
+ "step": 4900
303
+ },
304
+ {
305
+ "epoch": 3.91,
306
+ "learning_rate": 0.0001,
307
+ "loss": 4.3892,
308
+ "step": 5000
309
+ },
310
+ {
311
+ "epoch": 3.98,
312
+ "learning_rate": 0.0001,
313
+ "loss": 4.3886,
314
+ "step": 5100
315
+ },
316
+ {
317
+ "epoch": 4.06,
318
+ "learning_rate": 0.0001,
319
+ "loss": 4.3915,
320
+ "step": 5200
321
+ },
322
+ {
323
+ "epoch": 4.14,
324
+ "learning_rate": 0.0001,
325
+ "loss": 4.3706,
326
+ "step": 5300
327
+ },
328
+ {
329
+ "epoch": 4.22,
330
+ "learning_rate": 0.0001,
331
+ "loss": 4.3528,
332
+ "step": 5400
333
+ },
334
+ {
335
+ "epoch": 4.3,
336
+ "learning_rate": 0.0001,
337
+ "loss": 4.3272,
338
+ "step": 5500
339
+ },
340
+ {
341
+ "epoch": 4.38,
342
+ "learning_rate": 0.0001,
343
+ "loss": 4.3679,
344
+ "step": 5600
345
+ },
346
+ {
347
+ "epoch": 4.45,
348
+ "learning_rate": 0.0001,
349
+ "loss": 4.3608,
350
+ "step": 5700
351
+ },
352
+ {
353
+ "epoch": 4.53,
354
+ "learning_rate": 0.0001,
355
+ "loss": 4.3262,
356
+ "step": 5800
357
+ },
358
+ {
359
+ "epoch": 4.61,
360
+ "learning_rate": 0.0001,
361
+ "loss": 4.3097,
362
+ "step": 5900
363
+ },
364
+ {
365
+ "epoch": 4.69,
366
+ "learning_rate": 0.0001,
367
+ "loss": 4.2973,
368
+ "step": 6000
369
+ },
370
+ {
371
+ "epoch": 4.77,
372
+ "learning_rate": 0.0001,
373
+ "loss": 4.2756,
374
+ "step": 6100
375
+ },
376
+ {
377
+ "epoch": 4.84,
378
+ "learning_rate": 0.0001,
379
+ "loss": 4.3174,
380
+ "step": 6200
381
+ },
382
+ {
383
+ "epoch": 4.92,
384
+ "learning_rate": 0.0001,
385
+ "loss": 4.2786,
386
+ "step": 6300
387
+ },
388
+ {
389
+ "epoch": 5.0,
390
+ "learning_rate": 0.0001,
391
+ "loss": 4.2484,
392
+ "step": 6400
393
+ },
394
+ {
395
+ "epoch": 5.08,
396
+ "learning_rate": 0.0001,
397
+ "loss": 4.2634,
398
+ "step": 6500
399
+ },
400
+ {
401
+ "epoch": 5.16,
402
+ "learning_rate": 0.0001,
403
+ "loss": 4.228,
404
+ "step": 6600
405
+ },
406
+ {
407
+ "epoch": 5.23,
408
+ "learning_rate": 0.0001,
409
+ "loss": 4.2745,
410
+ "step": 6700
411
+ },
412
+ {
413
+ "epoch": 5.31,
414
+ "learning_rate": 0.0001,
415
+ "loss": 4.2297,
416
+ "step": 6800
417
+ },
418
+ {
419
+ "epoch": 5.39,
420
+ "learning_rate": 0.0001,
421
+ "loss": 4.1989,
422
+ "step": 6900
423
+ },
424
+ {
425
+ "epoch": 5.47,
426
+ "learning_rate": 0.0001,
427
+ "loss": 4.2074,
428
+ "step": 7000
429
+ },
430
+ {
431
+ "epoch": 5.55,
432
+ "learning_rate": 0.0001,
433
+ "loss": 4.201,
434
+ "step": 7100
435
+ },
436
+ {
437
+ "epoch": 5.62,
438
+ "learning_rate": 0.0001,
439
+ "loss": 4.2109,
440
+ "step": 7200
441
+ },
442
+ {
443
+ "epoch": 5.7,
444
+ "learning_rate": 0.0001,
445
+ "loss": 4.1945,
446
+ "step": 7300
447
+ },
448
+ {
449
+ "epoch": 5.78,
450
+ "learning_rate": 0.0001,
451
+ "loss": 4.1744,
452
+ "step": 7400
453
+ },
454
+ {
455
+ "epoch": 5.86,
456
+ "learning_rate": 0.0001,
457
+ "loss": 4.1741,
458
+ "step": 7500
459
+ },
460
+ {
461
+ "epoch": 5.94,
462
+ "learning_rate": 0.0001,
463
+ "loss": 4.19,
464
+ "step": 7600
465
+ },
466
+ {
467
+ "epoch": 6.02,
468
+ "learning_rate": 0.0001,
469
+ "loss": 4.1612,
470
+ "step": 7700
471
+ },
472
+ {
473
+ "epoch": 6.09,
474
+ "learning_rate": 0.0001,
475
+ "loss": 4.1559,
476
+ "step": 7800
477
+ },
478
+ {
479
+ "epoch": 6.17,
480
+ "learning_rate": 0.0001,
481
+ "loss": 4.1378,
482
+ "step": 7900
483
+ },
484
+ {
485
+ "epoch": 6.25,
486
+ "learning_rate": 0.0001,
487
+ "loss": 4.144,
488
+ "step": 8000
489
+ },
490
+ {
491
+ "epoch": 6.33,
492
+ "learning_rate": 0.0001,
493
+ "loss": 4.1601,
494
+ "step": 8100
495
+ },
496
+ {
497
+ "epoch": 6.41,
498
+ "learning_rate": 0.0001,
499
+ "loss": 4.1339,
500
+ "step": 8200
501
+ },
502
+ {
503
+ "epoch": 6.48,
504
+ "learning_rate": 0.0001,
505
+ "loss": 4.13,
506
+ "step": 8300
507
+ },
508
+ {
509
+ "epoch": 6.56,
510
+ "learning_rate": 0.0001,
511
+ "loss": 4.0877,
512
+ "step": 8400
513
+ },
514
+ {
515
+ "epoch": 6.64,
516
+ "learning_rate": 0.0001,
517
+ "loss": 4.1022,
518
+ "step": 8500
519
+ },
520
+ {
521
+ "epoch": 6.72,
522
+ "learning_rate": 0.0001,
523
+ "loss": 4.1034,
524
+ "step": 8600
525
+ },
526
+ {
527
+ "epoch": 6.8,
528
+ "learning_rate": 0.0001,
529
+ "loss": 4.091,
530
+ "step": 8700
531
+ },
532
+ {
533
+ "epoch": 6.88,
534
+ "learning_rate": 0.0001,
535
+ "loss": 4.0914,
536
+ "step": 8800
537
+ },
538
+ {
539
+ "epoch": 6.95,
540
+ "learning_rate": 0.0001,
541
+ "loss": 4.0937,
542
+ "step": 8900
543
+ },
544
+ {
545
+ "epoch": 7.03,
546
+ "learning_rate": 0.0001,
547
+ "loss": 4.0897,
548
+ "step": 9000
549
+ },
550
+ {
551
+ "epoch": 7.11,
552
+ "learning_rate": 0.0001,
553
+ "loss": 4.0844,
554
+ "step": 9100
555
+ },
556
+ {
557
+ "epoch": 7.19,
558
+ "learning_rate": 0.0001,
559
+ "loss": 4.0574,
560
+ "step": 9200
561
+ },
562
+ {
563
+ "epoch": 7.27,
564
+ "learning_rate": 0.0001,
565
+ "loss": 4.0837,
566
+ "step": 9300
567
+ },
568
+ {
569
+ "epoch": 7.34,
570
+ "learning_rate": 0.0001,
571
+ "loss": 4.0766,
572
+ "step": 9400
573
+ },
574
+ {
575
+ "epoch": 7.42,
576
+ "learning_rate": 0.0001,
577
+ "loss": 4.066,
578
+ "step": 9500
579
+ },
580
+ {
581
+ "epoch": 7.5,
582
+ "learning_rate": 0.0001,
583
+ "loss": 3.9944,
584
+ "step": 9600
585
+ },
586
+ {
587
+ "epoch": 7.58,
588
+ "learning_rate": 0.0001,
589
+ "loss": 4.0126,
590
+ "step": 9700
591
+ },
592
+ {
593
+ "epoch": 7.66,
594
+ "learning_rate": 0.0001,
595
+ "loss": 4.0273,
596
+ "step": 9800
597
+ },
598
+ {
599
+ "epoch": 7.73,
600
+ "learning_rate": 0.0001,
601
+ "loss": 4.0706,
602
+ "step": 9900
603
+ },
604
+ {
605
+ "epoch": 7.81,
606
+ "learning_rate": 0.0001,
607
+ "loss": 4.0008,
608
+ "step": 10000
609
+ },
610
+ {
611
+ "epoch": 7.89,
612
+ "learning_rate": 0.0001,
613
+ "loss": 4.0323,
614
+ "step": 10100
615
+ },
616
+ {
617
+ "epoch": 7.97,
618
+ "learning_rate": 0.0001,
619
+ "loss": 3.9864,
620
+ "step": 10200
621
+ },
622
+ {
623
+ "epoch": 8.05,
624
+ "learning_rate": 0.0001,
625
+ "loss": 4.0265,
626
+ "step": 10300
627
+ },
628
+ {
629
+ "epoch": 8.12,
630
+ "learning_rate": 0.0001,
631
+ "loss": 3.9874,
632
+ "step": 10400
633
+ },
634
+ {
635
+ "epoch": 8.2,
636
+ "learning_rate": 0.0001,
637
+ "loss": 3.9839,
638
+ "step": 10500
639
+ },
640
+ {
641
+ "epoch": 8.28,
642
+ "learning_rate": 0.0001,
643
+ "loss": 3.9583,
644
+ "step": 10600
645
+ },
646
+ {
647
+ "epoch": 8.36,
648
+ "learning_rate": 0.0001,
649
+ "loss": 4.007,
650
+ "step": 10700
651
+ },
652
+ {
653
+ "epoch": 8.44,
654
+ "learning_rate": 0.0001,
655
+ "loss": 3.9567,
656
+ "step": 10800
657
+ },
658
+ {
659
+ "epoch": 8.52,
660
+ "learning_rate": 0.0001,
661
+ "loss": 3.9888,
662
+ "step": 10900
663
+ },
664
+ {
665
+ "epoch": 8.59,
666
+ "learning_rate": 0.0001,
667
+ "loss": 3.9993,
668
+ "step": 11000
669
+ },
670
+ {
671
+ "epoch": 8.67,
672
+ "learning_rate": 0.0001,
673
+ "loss": 3.9914,
674
+ "step": 11100
675
+ },
676
+ {
677
+ "epoch": 8.75,
678
+ "learning_rate": 0.0001,
679
+ "loss": 4.0094,
680
+ "step": 11200
681
+ },
682
+ {
683
+ "epoch": 8.83,
684
+ "learning_rate": 0.0001,
685
+ "loss": 3.9338,
686
+ "step": 11300
687
+ },
688
+ {
689
+ "epoch": 8.91,
690
+ "learning_rate": 0.0001,
691
+ "loss": 3.9472,
692
+ "step": 11400
693
+ },
694
+ {
695
+ "epoch": 8.98,
696
+ "learning_rate": 0.0001,
697
+ "loss": 3.9595,
698
+ "step": 11500
699
+ },
700
+ {
701
+ "epoch": 9.06,
702
+ "learning_rate": 0.0001,
703
+ "loss": 3.9545,
704
+ "step": 11600
705
+ },
706
+ {
707
+ "epoch": 9.14,
708
+ "learning_rate": 0.0001,
709
+ "loss": 3.993,
710
+ "step": 11700
711
+ },
712
+ {
713
+ "epoch": 9.22,
714
+ "learning_rate": 0.0001,
715
+ "loss": 3.895,
716
+ "step": 11800
717
+ },
718
+ {
719
+ "epoch": 9.3,
720
+ "learning_rate": 0.0001,
721
+ "loss": 3.9527,
722
+ "step": 11900
723
+ },
724
+ {
725
+ "epoch": 9.38,
726
+ "learning_rate": 0.0001,
727
+ "loss": 3.9205,
728
+ "step": 12000
729
+ },
730
+ {
731
+ "epoch": 9.45,
732
+ "learning_rate": 0.0001,
733
+ "loss": 3.914,
734
+ "step": 12100
735
+ },
736
+ {
737
+ "epoch": 9.53,
738
+ "learning_rate": 0.0001,
739
+ "loss": 3.8872,
740
+ "step": 12200
741
+ },
742
+ {
743
+ "epoch": 9.61,
744
+ "learning_rate": 0.0001,
745
+ "loss": 3.9171,
746
+ "step": 12300
747
+ },
748
+ {
749
+ "epoch": 9.69,
750
+ "learning_rate": 0.0001,
751
+ "loss": 3.9148,
752
+ "step": 12400
753
+ }
754
+ ],
755
+ "max_steps": 12800,
756
+ "num_train_epochs": 10,
757
+ "total_flos": 9336653187287040.0,
758
+ "trial_name": null,
759
+ "trial_params": null
760
+ }
checkpoint-12480/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e683e2614c4dcf1ee5d4bfe9622ed2cfdbf3faf9319855e4a9b393368f21e44
3
+ size 3899
checkpoint-12510/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: fp4
15
+ - bnb_4bit_use_double_quant: False
16
+ - bnb_4bit_compute_dtype: float32
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.5.0.dev0
checkpoint-12510/adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "models/TinyStories-33M",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 8,
11
+ "lora_dropout": 0.0,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 16,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
checkpoint-12510/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b948a1d9a656951f8d37c2a23f3fef23ec4359f73d94eb2c5083218d521774ba
3
+ size 792353
checkpoint-12510/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c4677e1bd62cddb8d9dda555f90a9ba7f55b880e4cba14ceac1778354d9796f
3
+ size 1586873
checkpoint-12510/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8bb3860d98363e44bc4704fb9458deec927fd96997d1d09c80639070db959d22
3
+ size 14575
checkpoint-12510/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:226c5739a6beb840fe019bb17102960eab6abdb587f0b62865030f81daa87392
3
+ size 627
checkpoint-12510/trainer_state.json ADDED
@@ -0,0 +1,766 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 9.7734375,
5
+ "global_step": 12510,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.08,
12
+ "learning_rate": 0.0001,
13
+ "loss": 10.386,
14
+ "step": 100
15
+ },
16
+ {
17
+ "epoch": 0.16,
18
+ "learning_rate": 0.0001,
19
+ "loss": 7.4061,
20
+ "step": 200
21
+ },
22
+ {
23
+ "epoch": 0.23,
24
+ "learning_rate": 0.0001,
25
+ "loss": 6.4112,
26
+ "step": 300
27
+ },
28
+ {
29
+ "epoch": 0.31,
30
+ "learning_rate": 0.0001,
31
+ "loss": 6.091,
32
+ "step": 400
33
+ },
34
+ {
35
+ "epoch": 0.39,
36
+ "learning_rate": 0.0001,
37
+ "loss": 5.8571,
38
+ "step": 500
39
+ },
40
+ {
41
+ "epoch": 0.47,
42
+ "learning_rate": 0.0001,
43
+ "loss": 5.7144,
44
+ "step": 600
45
+ },
46
+ {
47
+ "epoch": 0.55,
48
+ "learning_rate": 0.0001,
49
+ "loss": 5.6228,
50
+ "step": 700
51
+ },
52
+ {
53
+ "epoch": 0.62,
54
+ "learning_rate": 0.0001,
55
+ "loss": 5.546,
56
+ "step": 800
57
+ },
58
+ {
59
+ "epoch": 0.7,
60
+ "learning_rate": 0.0001,
61
+ "loss": 5.4904,
62
+ "step": 900
63
+ },
64
+ {
65
+ "epoch": 0.78,
66
+ "learning_rate": 0.0001,
67
+ "loss": 5.4413,
68
+ "step": 1000
69
+ },
70
+ {
71
+ "epoch": 0.86,
72
+ "learning_rate": 0.0001,
73
+ "loss": 5.3603,
74
+ "step": 1100
75
+ },
76
+ {
77
+ "epoch": 0.94,
78
+ "learning_rate": 0.0001,
79
+ "loss": 5.3346,
80
+ "step": 1200
81
+ },
82
+ {
83
+ "epoch": 1.02,
84
+ "learning_rate": 0.0001,
85
+ "loss": 5.2744,
86
+ "step": 1300
87
+ },
88
+ {
89
+ "epoch": 1.09,
90
+ "learning_rate": 0.0001,
91
+ "loss": 5.231,
92
+ "step": 1400
93
+ },
94
+ {
95
+ "epoch": 1.17,
96
+ "learning_rate": 0.0001,
97
+ "loss": 5.1782,
98
+ "step": 1500
99
+ },
100
+ {
101
+ "epoch": 1.25,
102
+ "learning_rate": 0.0001,
103
+ "loss": 5.1613,
104
+ "step": 1600
105
+ },
106
+ {
107
+ "epoch": 1.33,
108
+ "learning_rate": 0.0001,
109
+ "loss": 5.1091,
110
+ "step": 1700
111
+ },
112
+ {
113
+ "epoch": 1.41,
114
+ "learning_rate": 0.0001,
115
+ "loss": 5.0939,
116
+ "step": 1800
117
+ },
118
+ {
119
+ "epoch": 1.48,
120
+ "learning_rate": 0.0001,
121
+ "loss": 5.0667,
122
+ "step": 1900
123
+ },
124
+ {
125
+ "epoch": 1.56,
126
+ "learning_rate": 0.0001,
127
+ "loss": 5.0054,
128
+ "step": 2000
129
+ },
130
+ {
131
+ "epoch": 1.64,
132
+ "learning_rate": 0.0001,
133
+ "loss": 4.96,
134
+ "step": 2100
135
+ },
136
+ {
137
+ "epoch": 1.72,
138
+ "learning_rate": 0.0001,
139
+ "loss": 4.9301,
140
+ "step": 2200
141
+ },
142
+ {
143
+ "epoch": 1.8,
144
+ "learning_rate": 0.0001,
145
+ "loss": 4.9302,
146
+ "step": 2300
147
+ },
148
+ {
149
+ "epoch": 1.88,
150
+ "learning_rate": 0.0001,
151
+ "loss": 4.9129,
152
+ "step": 2400
153
+ },
154
+ {
155
+ "epoch": 1.95,
156
+ "learning_rate": 0.0001,
157
+ "loss": 4.8627,
158
+ "step": 2500
159
+ },
160
+ {
161
+ "epoch": 2.03,
162
+ "learning_rate": 0.0001,
163
+ "loss": 4.8336,
164
+ "step": 2600
165
+ },
166
+ {
167
+ "epoch": 2.11,
168
+ "learning_rate": 0.0001,
169
+ "loss": 4.8018,
170
+ "step": 2700
171
+ },
172
+ {
173
+ "epoch": 2.19,
174
+ "learning_rate": 0.0001,
175
+ "loss": 4.7479,
176
+ "step": 2800
177
+ },
178
+ {
179
+ "epoch": 2.27,
180
+ "learning_rate": 0.0001,
181
+ "loss": 4.7572,
182
+ "step": 2900
183
+ },
184
+ {
185
+ "epoch": 2.34,
186
+ "learning_rate": 0.0001,
187
+ "loss": 4.759,
188
+ "step": 3000
189
+ },
190
+ {
191
+ "epoch": 2.42,
192
+ "learning_rate": 0.0001,
193
+ "loss": 4.7441,
194
+ "step": 3100
195
+ },
196
+ {
197
+ "epoch": 2.5,
198
+ "learning_rate": 0.0001,
199
+ "loss": 4.7195,
200
+ "step": 3200
201
+ },
202
+ {
203
+ "epoch": 2.58,
204
+ "learning_rate": 0.0001,
205
+ "loss": 4.6821,
206
+ "step": 3300
207
+ },
208
+ {
209
+ "epoch": 2.66,
210
+ "learning_rate": 0.0001,
211
+ "loss": 4.6587,
212
+ "step": 3400
213
+ },
214
+ {
215
+ "epoch": 2.73,
216
+ "learning_rate": 0.0001,
217
+ "loss": 4.6228,
218
+ "step": 3500
219
+ },
220
+ {
221
+ "epoch": 2.81,
222
+ "learning_rate": 0.0001,
223
+ "loss": 4.6248,
224
+ "step": 3600
225
+ },
226
+ {
227
+ "epoch": 2.89,
228
+ "learning_rate": 0.0001,
229
+ "loss": 4.5801,
230
+ "step": 3700
231
+ },
232
+ {
233
+ "epoch": 2.97,
234
+ "learning_rate": 0.0001,
235
+ "loss": 4.5934,
236
+ "step": 3800
237
+ },
238
+ {
239
+ "epoch": 3.05,
240
+ "learning_rate": 0.0001,
241
+ "loss": 4.5781,
242
+ "step": 3900
243
+ },
244
+ {
245
+ "epoch": 3.12,
246
+ "learning_rate": 0.0001,
247
+ "loss": 4.5763,
248
+ "step": 4000
249
+ },
250
+ {
251
+ "epoch": 3.2,
252
+ "learning_rate": 0.0001,
253
+ "loss": 4.5598,
254
+ "step": 4100
255
+ },
256
+ {
257
+ "epoch": 3.28,
258
+ "learning_rate": 0.0001,
259
+ "loss": 4.4796,
260
+ "step": 4200
261
+ },
262
+ {
263
+ "epoch": 3.36,
264
+ "learning_rate": 0.0001,
265
+ "loss": 4.5204,
266
+ "step": 4300
267
+ },
268
+ {
269
+ "epoch": 3.44,
270
+ "learning_rate": 0.0001,
271
+ "loss": 4.4858,
272
+ "step": 4400
273
+ },
274
+ {
275
+ "epoch": 3.52,
276
+ "learning_rate": 0.0001,
277
+ "loss": 4.4882,
278
+ "step": 4500
279
+ },
280
+ {
281
+ "epoch": 3.59,
282
+ "learning_rate": 0.0001,
283
+ "loss": 4.4852,
284
+ "step": 4600
285
+ },
286
+ {
287
+ "epoch": 3.67,
288
+ "learning_rate": 0.0001,
289
+ "loss": 4.4302,
290
+ "step": 4700
291
+ },
292
+ {
293
+ "epoch": 3.75,
294
+ "learning_rate": 0.0001,
295
+ "loss": 4.4267,
296
+ "step": 4800
297
+ },
298
+ {
299
+ "epoch": 3.83,
300
+ "learning_rate": 0.0001,
301
+ "loss": 4.4298,
302
+ "step": 4900
303
+ },
304
+ {
305
+ "epoch": 3.91,
306
+ "learning_rate": 0.0001,
307
+ "loss": 4.3892,
308
+ "step": 5000
309
+ },
310
+ {
311
+ "epoch": 3.98,
312
+ "learning_rate": 0.0001,
313
+ "loss": 4.3886,
314
+ "step": 5100
315
+ },
316
+ {
317
+ "epoch": 4.06,
318
+ "learning_rate": 0.0001,
319
+ "loss": 4.3915,
320
+ "step": 5200
321
+ },
322
+ {
323
+ "epoch": 4.14,
324
+ "learning_rate": 0.0001,
325
+ "loss": 4.3706,
326
+ "step": 5300
327
+ },
328
+ {
329
+ "epoch": 4.22,
330
+ "learning_rate": 0.0001,
331
+ "loss": 4.3528,
332
+ "step": 5400
333
+ },
334
+ {
335
+ "epoch": 4.3,
336
+ "learning_rate": 0.0001,
337
+ "loss": 4.3272,
338
+ "step": 5500
339
+ },
340
+ {
341
+ "epoch": 4.38,
342
+ "learning_rate": 0.0001,
343
+ "loss": 4.3679,
344
+ "step": 5600
345
+ },
346
+ {
347
+ "epoch": 4.45,
348
+ "learning_rate": 0.0001,
349
+ "loss": 4.3608,
350
+ "step": 5700
351
+ },
352
+ {
353
+ "epoch": 4.53,
354
+ "learning_rate": 0.0001,
355
+ "loss": 4.3262,
356
+ "step": 5800
357
+ },
358
+ {
359
+ "epoch": 4.61,
360
+ "learning_rate": 0.0001,
361
+ "loss": 4.3097,
362
+ "step": 5900
363
+ },
364
+ {
365
+ "epoch": 4.69,
366
+ "learning_rate": 0.0001,
367
+ "loss": 4.2973,
368
+ "step": 6000
369
+ },
370
+ {
371
+ "epoch": 4.77,
372
+ "learning_rate": 0.0001,
373
+ "loss": 4.2756,
374
+ "step": 6100
375
+ },
376
+ {
377
+ "epoch": 4.84,
378
+ "learning_rate": 0.0001,
379
+ "loss": 4.3174,
380
+ "step": 6200
381
+ },
382
+ {
383
+ "epoch": 4.92,
384
+ "learning_rate": 0.0001,
385
+ "loss": 4.2786,
386
+ "step": 6300
387
+ },
388
+ {
389
+ "epoch": 5.0,
390
+ "learning_rate": 0.0001,
391
+ "loss": 4.2484,
392
+ "step": 6400
393
+ },
394
+ {
395
+ "epoch": 5.08,
396
+ "learning_rate": 0.0001,
397
+ "loss": 4.2634,
398
+ "step": 6500
399
+ },
400
+ {
401
+ "epoch": 5.16,
402
+ "learning_rate": 0.0001,
403
+ "loss": 4.228,
404
+ "step": 6600
405
+ },
406
+ {
407
+ "epoch": 5.23,
408
+ "learning_rate": 0.0001,
409
+ "loss": 4.2745,
410
+ "step": 6700
411
+ },
412
+ {
413
+ "epoch": 5.31,
414
+ "learning_rate": 0.0001,
415
+ "loss": 4.2297,
416
+ "step": 6800
417
+ },
418
+ {
419
+ "epoch": 5.39,
420
+ "learning_rate": 0.0001,
421
+ "loss": 4.1989,
422
+ "step": 6900
423
+ },
424
+ {
425
+ "epoch": 5.47,
426
+ "learning_rate": 0.0001,
427
+ "loss": 4.2074,
428
+ "step": 7000
429
+ },
430
+ {
431
+ "epoch": 5.55,
432
+ "learning_rate": 0.0001,
433
+ "loss": 4.201,
434
+ "step": 7100
435
+ },
436
+ {
437
+ "epoch": 5.62,
438
+ "learning_rate": 0.0001,
439
+ "loss": 4.2109,
440
+ "step": 7200
441
+ },
442
+ {
443
+ "epoch": 5.7,
444
+ "learning_rate": 0.0001,
445
+ "loss": 4.1945,
446
+ "step": 7300
447
+ },
448
+ {
449
+ "epoch": 5.78,
450
+ "learning_rate": 0.0001,
451
+ "loss": 4.1744,
452
+ "step": 7400
453
+ },
454
+ {
455
+ "epoch": 5.86,
456
+ "learning_rate": 0.0001,
457
+ "loss": 4.1741,
458
+ "step": 7500
459
+ },
460
+ {
461
+ "epoch": 5.94,
462
+ "learning_rate": 0.0001,
463
+ "loss": 4.19,
464
+ "step": 7600
465
+ },
466
+ {
467
+ "epoch": 6.02,
468
+ "learning_rate": 0.0001,
469
+ "loss": 4.1612,
470
+ "step": 7700
471
+ },
472
+ {
473
+ "epoch": 6.09,
474
+ "learning_rate": 0.0001,
475
+ "loss": 4.1559,
476
+ "step": 7800
477
+ },
478
+ {
479
+ "epoch": 6.17,
480
+ "learning_rate": 0.0001,
481
+ "loss": 4.1378,
482
+ "step": 7900
483
+ },
484
+ {
485
+ "epoch": 6.25,
486
+ "learning_rate": 0.0001,
487
+ "loss": 4.144,
488
+ "step": 8000
489
+ },
490
+ {
491
+ "epoch": 6.33,
492
+ "learning_rate": 0.0001,
493
+ "loss": 4.1601,
494
+ "step": 8100
495
+ },
496
+ {
497
+ "epoch": 6.41,
498
+ "learning_rate": 0.0001,
499
+ "loss": 4.1339,
500
+ "step": 8200
501
+ },
502
+ {
503
+ "epoch": 6.48,
504
+ "learning_rate": 0.0001,
505
+ "loss": 4.13,
506
+ "step": 8300
507
+ },
508
+ {
509
+ "epoch": 6.56,
510
+ "learning_rate": 0.0001,
511
+ "loss": 4.0877,
512
+ "step": 8400
513
+ },
514
+ {
515
+ "epoch": 6.64,
516
+ "learning_rate": 0.0001,
517
+ "loss": 4.1022,
518
+ "step": 8500
519
+ },
520
+ {
521
+ "epoch": 6.72,
522
+ "learning_rate": 0.0001,
523
+ "loss": 4.1034,
524
+ "step": 8600
525
+ },
526
+ {
527
+ "epoch": 6.8,
528
+ "learning_rate": 0.0001,
529
+ "loss": 4.091,
530
+ "step": 8700
531
+ },
532
+ {
533
+ "epoch": 6.88,
534
+ "learning_rate": 0.0001,
535
+ "loss": 4.0914,
536
+ "step": 8800
537
+ },
538
+ {
539
+ "epoch": 6.95,
540
+ "learning_rate": 0.0001,
541
+ "loss": 4.0937,
542
+ "step": 8900
543
+ },
544
+ {
545
+ "epoch": 7.03,
546
+ "learning_rate": 0.0001,
547
+ "loss": 4.0897,
548
+ "step": 9000
549
+ },
550
+ {
551
+ "epoch": 7.11,
552
+ "learning_rate": 0.0001,
553
+ "loss": 4.0844,
554
+ "step": 9100
555
+ },
556
+ {
557
+ "epoch": 7.19,
558
+ "learning_rate": 0.0001,
559
+ "loss": 4.0574,
560
+ "step": 9200
561
+ },
562
+ {
563
+ "epoch": 7.27,
564
+ "learning_rate": 0.0001,
565
+ "loss": 4.0837,
566
+ "step": 9300
567
+ },
568
+ {
569
+ "epoch": 7.34,
570
+ "learning_rate": 0.0001,
571
+ "loss": 4.0766,
572
+ "step": 9400
573
+ },
574
+ {
575
+ "epoch": 7.42,
576
+ "learning_rate": 0.0001,
577
+ "loss": 4.066,
578
+ "step": 9500
579
+ },
580
+ {
581
+ "epoch": 7.5,
582
+ "learning_rate": 0.0001,
583
+ "loss": 3.9944,
584
+ "step": 9600
585
+ },
586
+ {
587
+ "epoch": 7.58,
588
+ "learning_rate": 0.0001,
589
+ "loss": 4.0126,
590
+ "step": 9700
591
+ },
592
+ {
593
+ "epoch": 7.66,
594
+ "learning_rate": 0.0001,
595
+ "loss": 4.0273,
596
+ "step": 9800
597
+ },
598
+ {
599
+ "epoch": 7.73,
600
+ "learning_rate": 0.0001,
601
+ "loss": 4.0706,
602
+ "step": 9900
603
+ },
604
+ {
605
+ "epoch": 7.81,
606
+ "learning_rate": 0.0001,
607
+ "loss": 4.0008,
608
+ "step": 10000
609
+ },
610
+ {
611
+ "epoch": 7.89,
612
+ "learning_rate": 0.0001,
613
+ "loss": 4.0323,
614
+ "step": 10100
615
+ },
616
+ {
617
+ "epoch": 7.97,
618
+ "learning_rate": 0.0001,
619
+ "loss": 3.9864,
620
+ "step": 10200
621
+ },
622
+ {
623
+ "epoch": 8.05,
624
+ "learning_rate": 0.0001,
625
+ "loss": 4.0265,
626
+ "step": 10300
627
+ },
628
+ {
629
+ "epoch": 8.12,
630
+ "learning_rate": 0.0001,
631
+ "loss": 3.9874,
632
+ "step": 10400
633
+ },
634
+ {
635
+ "epoch": 8.2,
636
+ "learning_rate": 0.0001,
637
+ "loss": 3.9839,
638
+ "step": 10500
639
+ },
640
+ {
641
+ "epoch": 8.28,
642
+ "learning_rate": 0.0001,
643
+ "loss": 3.9583,
644
+ "step": 10600
645
+ },
646
+ {
647
+ "epoch": 8.36,
648
+ "learning_rate": 0.0001,
649
+ "loss": 4.007,
650
+ "step": 10700
651
+ },
652
+ {
653
+ "epoch": 8.44,
654
+ "learning_rate": 0.0001,
655
+ "loss": 3.9567,
656
+ "step": 10800
657
+ },
658
+ {
659
+ "epoch": 8.52,
660
+ "learning_rate": 0.0001,
661
+ "loss": 3.9888,
662
+ "step": 10900
663
+ },
664
+ {
665
+ "epoch": 8.59,
666
+ "learning_rate": 0.0001,
667
+ "loss": 3.9993,
668
+ "step": 11000
669
+ },
670
+ {
671
+ "epoch": 8.67,
672
+ "learning_rate": 0.0001,
673
+ "loss": 3.9914,
674
+ "step": 11100
675
+ },
676
+ {
677
+ "epoch": 8.75,
678
+ "learning_rate": 0.0001,
679
+ "loss": 4.0094,
680
+ "step": 11200
681
+ },
682
+ {
683
+ "epoch": 8.83,
684
+ "learning_rate": 0.0001,
685
+ "loss": 3.9338,
686
+ "step": 11300
687
+ },
688
+ {
689
+ "epoch": 8.91,
690
+ "learning_rate": 0.0001,
691
+ "loss": 3.9472,
692
+ "step": 11400
693
+ },
694
+ {
695
+ "epoch": 8.98,
696
+ "learning_rate": 0.0001,
697
+ "loss": 3.9595,
698
+ "step": 11500
699
+ },
700
+ {
701
+ "epoch": 9.06,
702
+ "learning_rate": 0.0001,
703
+ "loss": 3.9545,
704
+ "step": 11600
705
+ },
706
+ {
707
+ "epoch": 9.14,
708
+ "learning_rate": 0.0001,
709
+ "loss": 3.993,
710
+ "step": 11700
711
+ },
712
+ {
713
+ "epoch": 9.22,
714
+ "learning_rate": 0.0001,
715
+ "loss": 3.895,
716
+ "step": 11800
717
+ },
718
+ {
719
+ "epoch": 9.3,
720
+ "learning_rate": 0.0001,
721
+ "loss": 3.9527,
722
+ "step": 11900
723
+ },
724
+ {
725
+ "epoch": 9.38,
726
+ "learning_rate": 0.0001,
727
+ "loss": 3.9205,
728
+ "step": 12000
729
+ },
730
+ {
731
+ "epoch": 9.45,
732
+ "learning_rate": 0.0001,
733
+ "loss": 3.914,
734
+ "step": 12100
735
+ },
736
+ {
737
+ "epoch": 9.53,
738
+ "learning_rate": 0.0001,
739
+ "loss": 3.8872,
740
+ "step": 12200
741
+ },
742
+ {
743
+ "epoch": 9.61,
744
+ "learning_rate": 0.0001,
745
+ "loss": 3.9171,
746
+ "step": 12300
747
+ },
748
+ {
749
+ "epoch": 9.69,
750
+ "learning_rate": 0.0001,
751
+ "loss": 3.9148,
752
+ "step": 12400
753
+ },
754
+ {
755
+ "epoch": 9.77,
756
+ "learning_rate": 0.0001,
757
+ "loss": 3.9181,
758
+ "step": 12500
759
+ }
760
+ ],
761
+ "max_steps": 12800,
762
+ "num_train_epochs": 10,
763
+ "total_flos": 9357159930071040.0,
764
+ "trial_name": null,
765
+ "trial_params": null
766
+ }