Model save
Browse files- README.md +2 -2
- all_results.json +6 -5
- config.json +1 -1
- generation_config.json +1 -1
- model-00001-of-00003.safetensors +1 -1
- model-00002-of-00003.safetensors +1 -1
- model-00003-of-00003.safetensors +1 -1
- runs/Jun30_15-54-56_n136-082-130/events.out.tfevents.1719734116.n136-082-130.3539205.0 +2 -2
- tokenizer.json +1 -0
- train_results.json +6 -5
- trainer_state.json +470 -458
- training_args.bin +2 -2
README.md
CHANGED
@@ -52,7 +52,7 @@ The following hyperparameters were used during training:
|
|
52 |
|
53 |
### Framework versions
|
54 |
|
55 |
-
- Transformers 4.
|
56 |
- Pytorch 2.1.2+cu118
|
57 |
- Datasets 2.16.1
|
58 |
-
- Tokenizers 0.
|
|
|
52 |
|
53 |
### Framework versions
|
54 |
|
55 |
+
- Transformers 4.41.1
|
56 |
- Pytorch 2.1.2+cu118
|
57 |
- Datasets 2.16.1
|
58 |
+
- Tokenizers 0.19.1
|
all_results.json
CHANGED
@@ -1,8 +1,9 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"
|
4 |
-
"
|
|
|
5 |
"train_samples": 52922,
|
6 |
-
"train_samples_per_second":
|
7 |
-
"train_steps_per_second": 0.
|
8 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 0.9987908101571947,
|
3 |
+
"total_flos": 0.0,
|
4 |
+
"train_loss": 0.9438655665076674,
|
5 |
+
"train_runtime": 7020.7263,
|
6 |
"train_samples": 52922,
|
7 |
+
"train_samples_per_second": 7.538,
|
8 |
+
"train_steps_per_second": 0.059
|
9 |
}
|
config.json
CHANGED
@@ -20,7 +20,7 @@
|
|
20 |
"sliding_window": 4096,
|
21 |
"tie_word_embeddings": false,
|
22 |
"torch_dtype": "bfloat16",
|
23 |
-
"transformers_version": "4.
|
24 |
"use_cache": false,
|
25 |
"vocab_size": 32000
|
26 |
}
|
|
|
20 |
"sliding_window": 4096,
|
21 |
"tie_word_embeddings": false,
|
22 |
"torch_dtype": "bfloat16",
|
23 |
+
"transformers_version": "4.41.1",
|
24 |
"use_cache": false,
|
25 |
"vocab_size": 32000
|
26 |
}
|
generation_config.json
CHANGED
@@ -2,5 +2,5 @@
|
|
2 |
"_from_model_config": true,
|
3 |
"bos_token_id": 1,
|
4 |
"eos_token_id": 2,
|
5 |
-
"transformers_version": "4.
|
6 |
}
|
|
|
2 |
"_from_model_config": true,
|
3 |
"bos_token_id": 1,
|
4 |
"eos_token_id": 2,
|
5 |
+
"transformers_version": "4.41.1"
|
6 |
}
|
model-00001-of-00003.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 4943162336
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:999f423f854e13fc506453d3f2abf8f89fa836065069fe6dc0288669f4e509e9
|
3 |
size 4943162336
|
model-00002-of-00003.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 4999819336
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b06e8c9d65bf6da14fcfbf182b621f376d131e989115189e52c8989e9918d89e
|
3 |
size 4999819336
|
model-00003-of-00003.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 4540516344
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f238dd9434885914394ed70efc6f48253db2b43b50dee3f8e6d21d12e952102a
|
3 |
size 4540516344
|
runs/Jun30_15-54-56_n136-082-130/events.out.tfevents.1719734116.n136-082-130.3539205.0
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9d65198d6efd08eae5bf1772fbfb49100799bd12e4569bb1c2a51a76f12d0b99
|
3 |
+
size 34199
|
tokenizer.json
CHANGED
@@ -134,6 +134,7 @@
|
|
134 |
"end_of_word_suffix": null,
|
135 |
"fuse_unk": true,
|
136 |
"byte_fallback": true,
|
|
|
137 |
"vocab": {
|
138 |
"<unk>": 0,
|
139 |
"<s>": 1,
|
|
|
134 |
"end_of_word_suffix": null,
|
135 |
"fuse_unk": true,
|
136 |
"byte_fallback": true,
|
137 |
+
"ignore_merges": false,
|
138 |
"vocab": {
|
139 |
"<unk>": 0,
|
140 |
"<s>": 1,
|
train_results.json
CHANGED
@@ -1,8 +1,9 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"
|
4 |
-
"
|
|
|
5 |
"train_samples": 52922,
|
6 |
-
"train_samples_per_second":
|
7 |
-
"train_steps_per_second": 0.
|
8 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 0.9987908101571947,
|
3 |
+
"total_flos": 0.0,
|
4 |
+
"train_loss": 0.9438655665076674,
|
5 |
+
"train_runtime": 7020.7263,
|
6 |
"train_samples": 52922,
|
7 |
+
"train_samples_per_second": 7.538,
|
8 |
+
"train_steps_per_second": 0.059
|
9 |
}
|
trainer_state.json
CHANGED
@@ -9,8 +9,8 @@
|
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
-
"epoch": 0.
|
13 |
-
"grad_norm":
|
14 |
"learning_rate": 9.523809523809522e-09,
|
15 |
"logits/chosen": -2.7005977630615234,
|
16 |
"logits/rejected": -2.6288318634033203,
|
@@ -24,628 +24,628 @@
|
|
24 |
"step": 1
|
25 |
},
|
26 |
{
|
27 |
-
"epoch": 0.
|
28 |
-
"grad_norm":
|
29 |
"learning_rate": 9.523809523809523e-08,
|
30 |
-
"logits/chosen": -2.
|
31 |
-
"logits/rejected": -2.
|
32 |
-
"logps/chosen": -0.
|
33 |
-
"logps/rejected": -0.
|
34 |
-
"loss": 0.
|
35 |
-
"rewards/accuracies": 0.
|
36 |
-
"rewards/chosen":
|
37 |
-
"rewards/margins": 0.
|
38 |
-
"rewards/rejected": -0.
|
39 |
"step": 10
|
40 |
},
|
41 |
{
|
42 |
-
"epoch": 0.
|
43 |
-
"grad_norm":
|
44 |
"learning_rate": 1.9047619047619045e-07,
|
45 |
-
"logits/chosen": -2.
|
46 |
-
"logits/rejected": -2.
|
47 |
-
"logps/chosen": -
|
48 |
-
"logps/rejected": -1.
|
49 |
-
"loss": 0.
|
50 |
-
"rewards/accuracies": 0.
|
51 |
-
"rewards/chosen": -0.
|
52 |
-
"rewards/margins": 0.
|
53 |
-
"rewards/rejected": -0.
|
54 |
"step": 20
|
55 |
},
|
56 |
{
|
57 |
-
"epoch": 0.
|
58 |
-
"grad_norm":
|
59 |
"learning_rate": 2.857142857142857e-07,
|
60 |
-
"logits/chosen": -2.
|
61 |
-
"logits/rejected": -2.
|
62 |
-
"logps/chosen": -
|
63 |
-
"logps/rejected": -
|
64 |
-
"loss": 0.
|
65 |
-
"rewards/accuracies": 0.
|
66 |
-
"rewards/chosen": -0.
|
67 |
-
"rewards/margins": 0.
|
68 |
-
"rewards/rejected": -
|
69 |
"step": 30
|
70 |
},
|
71 |
{
|
72 |
-
"epoch": 0.
|
73 |
-
"grad_norm":
|
74 |
"learning_rate": 3.809523809523809e-07,
|
75 |
-
"logits/chosen": -2.
|
76 |
-
"logits/rejected": -2.
|
77 |
-
"logps/chosen": -
|
78 |
-
"logps/rejected": -
|
79 |
-
"loss": 0.
|
80 |
-
"rewards/accuracies": 0.
|
81 |
-
"rewards/chosen": -
|
82 |
-
"rewards/margins":
|
83 |
-
"rewards/rejected": -
|
84 |
"step": 40
|
85 |
},
|
86 |
{
|
87 |
-
"epoch": 0.
|
88 |
-
"grad_norm":
|
89 |
"learning_rate": 3.995412608484087e-07,
|
90 |
-
"logits/chosen": -2.
|
91 |
-
"logits/rejected": -2.
|
92 |
-
"logps/chosen": -
|
93 |
-
"logps/rejected": -
|
94 |
-
"loss": 0.
|
95 |
-
"rewards/accuracies": 0.
|
96 |
-
"rewards/chosen":
|
97 |
-
"rewards/margins":
|
98 |
-
"rewards/rejected": -0.
|
99 |
"step": 50
|
100 |
},
|
101 |
{
|
102 |
-
"epoch": 0.
|
103 |
-
"grad_norm":
|
104 |
"learning_rate": 3.976812391485896e-07,
|
105 |
-
"logits/chosen": -2.
|
106 |
-
"logits/rejected": -2.
|
107 |
-
"logps/chosen": -
|
108 |
-
"logps/rejected": -
|
109 |
-
"loss":
|
110 |
-
"rewards/accuracies": 0.
|
111 |
-
"rewards/chosen": -
|
112 |
-
"rewards/margins":
|
113 |
-
"rewards/rejected": -
|
114 |
"step": 60
|
115 |
},
|
116 |
{
|
117 |
-
"epoch": 0.
|
118 |
-
"grad_norm":
|
119 |
"learning_rate": 3.9440458281608213e-07,
|
120 |
-
"logits/chosen": -2.
|
121 |
-
"logits/rejected": -2.
|
122 |
-
"logps/chosen": -
|
123 |
-
"logps/rejected": -
|
124 |
-
"loss": 0.
|
125 |
-
"rewards/accuracies": 0.
|
126 |
-
"rewards/chosen":
|
127 |
-
"rewards/margins":
|
128 |
-
"rewards/rejected": -1.
|
129 |
"step": 70
|
130 |
},
|
131 |
{
|
132 |
-
"epoch": 0.
|
133 |
-
"grad_norm":
|
134 |
"learning_rate": 3.897347732134074e-07,
|
135 |
-
"logits/chosen": -2.
|
136 |
-
"logits/rejected": -2.
|
137 |
-
"logps/chosen": -
|
138 |
-
"logps/rejected": -1.
|
139 |
-
"loss":
|
140 |
-
"rewards/accuracies": 0.
|
141 |
-
"rewards/chosen": -
|
142 |
-
"rewards/margins":
|
143 |
-
"rewards/rejected": -
|
144 |
"step": 80
|
145 |
},
|
146 |
{
|
147 |
-
"epoch": 0.
|
148 |
-
"grad_norm":
|
149 |
"learning_rate": 3.8370527539794614e-07,
|
150 |
-
"logits/chosen": -2.
|
151 |
-
"logits/rejected": -2.
|
152 |
-
"logps/chosen": -1.
|
153 |
-
"logps/rejected": -1.
|
154 |
-
"loss":
|
155 |
-
"rewards/accuracies": 0.
|
156 |
-
"rewards/chosen": -
|
157 |
-
"rewards/margins":
|
158 |
-
"rewards/rejected": -
|
159 |
"step": 90
|
160 |
},
|
161 |
{
|
162 |
-
"epoch": 0.
|
163 |
-
"grad_norm":
|
164 |
"learning_rate": 3.763592983027255e-07,
|
165 |
-
"logits/chosen": -2.
|
166 |
-
"logits/rejected": -2.
|
167 |
-
"logps/chosen": -
|
168 |
-
"logps/rejected": -
|
169 |
-
"loss":
|
170 |
-
"rewards/accuracies": 0.
|
171 |
-
"rewards/chosen": -
|
172 |
-
"rewards/margins":
|
173 |
-
"rewards/rejected": -
|
174 |
"step": 100
|
175 |
},
|
176 |
{
|
177 |
-
"epoch": 0.
|
178 |
-
"grad_norm":
|
179 |
"learning_rate": 3.6774948509008527e-07,
|
180 |
-
"logits/chosen": -
|
181 |
-
"logits/rejected": -2.
|
182 |
-
"logps/chosen": -
|
183 |
-
"logps/rejected": -
|
184 |
-
"loss": 0.
|
185 |
-
"rewards/accuracies": 0.
|
186 |
-
"rewards/chosen": -
|
187 |
-
"rewards/margins":
|
188 |
-
"rewards/rejected": -
|
189 |
"step": 110
|
190 |
},
|
191 |
{
|
192 |
-
"epoch": 0.
|
193 |
-
"grad_norm":
|
194 |
"learning_rate": 3.579375358972288e-07,
|
195 |
-
"logits/chosen": -
|
196 |
-
"logits/rejected": -2.
|
197 |
-
"logps/chosen": -
|
198 |
-
"logps/rejected": -1.
|
199 |
-
"loss": 0.
|
200 |
-
"rewards/accuracies": 0.
|
201 |
-
"rewards/chosen": -2.
|
202 |
-
"rewards/margins":
|
203 |
-
"rewards/rejected": -
|
204 |
"step": 120
|
205 |
},
|
206 |
{
|
207 |
-
"epoch": 0.
|
208 |
-
"grad_norm":
|
209 |
"learning_rate": 3.4699376567716156e-07,
|
210 |
-
"logits/chosen": -
|
211 |
-
"logits/rejected": -2.
|
212 |
-
"logps/chosen": -
|
213 |
-
"logps/rejected": -
|
214 |
-
"loss":
|
215 |
-
"rewards/accuracies": 0.
|
216 |
-
"rewards/chosen":
|
217 |
-
"rewards/margins":
|
218 |
-
"rewards/rejected": -
|
219 |
"step": 130
|
220 |
},
|
221 |
{
|
222 |
-
"epoch": 0.
|
223 |
-
"grad_norm":
|
224 |
"learning_rate": 3.349966003036421e-07,
|
225 |
-
"logits/chosen": -2.
|
226 |
-
"logits/rejected": -2.
|
227 |
-
"logps/chosen": -
|
228 |
-
"logps/rejected": -
|
229 |
-
"loss": 0.
|
230 |
-
"rewards/accuracies": 0.
|
231 |
-
"rewards/chosen": -1.
|
232 |
-
"rewards/margins":
|
233 |
-
"rewards/rejected": -
|
234 |
"step": 140
|
235 |
},
|
236 |
{
|
237 |
-
"epoch": 0.
|
238 |
-
"grad_norm":
|
239 |
"learning_rate": 3.220320145511884e-07,
|
240 |
-
"logits/chosen": -2.
|
241 |
-
"logits/rejected": -2.
|
242 |
-
"logps/chosen": -
|
243 |
-
"logps/rejected": -
|
244 |
-
"loss":
|
245 |
-
"rewards/accuracies": 0.
|
246 |
-
"rewards/chosen":
|
247 |
-
"rewards/margins":
|
248 |
-
"rewards/rejected": -3.
|
249 |
"step": 150
|
250 |
},
|
251 |
{
|
252 |
-
"epoch": 0.
|
253 |
-
"grad_norm":
|
254 |
"learning_rate": 3.0819291597771795e-07,
|
255 |
-
"logits/chosen": -2.
|
256 |
-
"logits/rejected": -2.
|
257 |
-
"logps/chosen": -
|
258 |
-
"logps/rejected": -
|
259 |
-
"loss":
|
260 |
-
"rewards/accuracies": 0.
|
261 |
-
"rewards/chosen": -
|
262 |
-
"rewards/margins":
|
263 |
-
"rewards/rejected": -
|
264 |
"step": 160
|
265 |
},
|
266 |
{
|
267 |
-
"epoch": 0.
|
268 |
-
"grad_norm":
|
269 |
"learning_rate": 2.9357847912507786e-07,
|
270 |
-
"logits/chosen": -2.
|
271 |
-
"logits/rejected": -2.
|
272 |
-
"logps/chosen": -
|
273 |
-
"logps/rejected": -
|
274 |
-
"loss":
|
275 |
-
"rewards/accuracies": 0.
|
276 |
-
"rewards/chosen": -
|
277 |
-
"rewards/margins":
|
278 |
-
"rewards/rejected": -
|
279 |
"step": 170
|
280 |
},
|
281 |
{
|
282 |
-
"epoch": 0.
|
283 |
-
"grad_norm":
|
284 |
"learning_rate": 2.7829343480875617e-07,
|
285 |
-
"logits/chosen": -2.
|
286 |
-
"logits/rejected": -2.
|
287 |
-
"logps/chosen": -
|
288 |
-
"logps/rejected": -
|
289 |
-
"loss":
|
290 |
-
"rewards/accuracies": 0.
|
291 |
-
"rewards/chosen":
|
292 |
-
"rewards/margins":
|
293 |
-
"rewards/rejected": -
|
294 |
"step": 180
|
295 |
},
|
296 |
{
|
297 |
-
"epoch": 0.
|
298 |
-
"grad_norm":
|
299 |
"learning_rate": 2.624473195899052e-07,
|
300 |
-
"logits/chosen": -2.
|
301 |
-
"logits/rejected": -2.
|
302 |
-
"logps/chosen": -
|
303 |
-
"logps/rejected": -
|
304 |
-
"loss":
|
305 |
-
"rewards/accuracies": 0.
|
306 |
-
"rewards/chosen": -
|
307 |
-
"rewards/margins":
|
308 |
-
"rewards/rejected": -
|
309 |
"step": 190
|
310 |
},
|
311 |
{
|
312 |
-
"epoch": 0.
|
313 |
-
"grad_norm":
|
314 |
"learning_rate": 2.4615369080815547e-07,
|
315 |
-
"logits/chosen": -2.
|
316 |
-
"logits/rejected": -2.
|
317 |
-
"logps/chosen": -
|
318 |
-
"logps/rejected": -
|
319 |
-
"loss":
|
320 |
-
"rewards/accuracies": 0.
|
321 |
-
"rewards/chosen":
|
322 |
-
"rewards/margins":
|
323 |
-
"rewards/rejected": -
|
324 |
"step": 200
|
325 |
},
|
326 |
{
|
327 |
-
"epoch": 0.
|
328 |
-
"grad_norm":
|
329 |
"learning_rate": 2.2952931280049625e-07,
|
330 |
-
"logits/chosen": -2.
|
331 |
-
"logits/rejected": -2.
|
332 |
-
"logps/chosen": -1.
|
333 |
-
"logps/rejected": -
|
334 |
-
"loss":
|
335 |
-
"rewards/accuracies": 0.
|
336 |
-
"rewards/chosen":
|
337 |
-
"rewards/margins":
|
338 |
-
"rewards/rejected": -
|
339 |
"step": 210
|
340 |
},
|
341 |
{
|
342 |
-
"epoch": 0.
|
343 |
-
"grad_norm":
|
344 |
"learning_rate": 2.1269332013798747e-07,
|
345 |
-
"logits/chosen": -2.
|
346 |
-
"logits/rejected": -2.
|
347 |
-
"logps/chosen": -
|
348 |
-
"logps/rejected": -
|
349 |
-
"loss":
|
350 |
-
"rewards/accuracies": 0.
|
351 |
-
"rewards/chosen":
|
352 |
-
"rewards/margins":
|
353 |
-
"rewards/rejected": -
|
354 |
"step": 220
|
355 |
},
|
356 |
{
|
357 |
-
"epoch": 0.
|
358 |
-
"grad_norm":
|
359 |
"learning_rate": 1.9576636387676436e-07,
|
360 |
-
"logits/chosen": -2.
|
361 |
-
"logits/rejected": -2.
|
362 |
-
"logps/chosen": -
|
363 |
-
"logps/rejected": -
|
364 |
-
"loss": 0.
|
365 |
-
"rewards/accuracies": 0.
|
366 |
-
"rewards/chosen": -1.
|
367 |
-
"rewards/margins":
|
368 |
-
"rewards/rejected": -
|
369 |
"step": 230
|
370 |
},
|
371 |
{
|
372 |
-
"epoch": 0.
|
373 |
-
"grad_norm":
|
374 |
"learning_rate": 1.7886974694151976e-07,
|
375 |
-
"logits/chosen": -2.
|
376 |
-
"logits/rejected": -2.
|
377 |
-
"logps/chosen": -
|
378 |
-
"logps/rejected": -
|
379 |
-
"loss":
|
380 |
-
"rewards/accuracies": 0.
|
381 |
-
"rewards/chosen": -
|
382 |
-
"rewards/margins":
|
383 |
-
"rewards/rejected": -
|
384 |
"step": 240
|
385 |
},
|
386 |
{
|
387 |
-
"epoch": 0.
|
388 |
-
"grad_norm":
|
389 |
"learning_rate": 1.6212455483752895e-07,
|
390 |
-
"logits/chosen": -2.
|
391 |
-
"logits/rejected": -2.
|
392 |
-
"logps/chosen": -
|
393 |
-
"logps/rejected": -
|
394 |
-
"loss":
|
395 |
-
"rewards/accuracies": 0.
|
396 |
-
"rewards/chosen":
|
397 |
-
"rewards/margins":
|
398 |
-
"rewards/rejected": -
|
399 |
"step": 250
|
400 |
},
|
401 |
{
|
402 |
-
"epoch": 0.
|
403 |
-
"grad_norm":
|
404 |
"learning_rate": 1.4565078792075733e-07,
|
405 |
-
"logits/chosen": -2.
|
406 |
-
"logits/rejected": -2.
|
407 |
-
"logps/chosen": -
|
408 |
-
"logps/rejected": -
|
409 |
-
"loss":
|
410 |
-
"rewards/accuracies": 0.
|
411 |
-
"rewards/chosen":
|
412 |
-
"rewards/margins":
|
413 |
-
"rewards/rejected": -
|
414 |
"step": 260
|
415 |
},
|
416 |
{
|
417 |
-
"epoch": 0.
|
418 |
-
"grad_norm":
|
419 |
"learning_rate": 1.295665014444281e-07,
|
420 |
-
"logits/chosen": -2.
|
421 |
-
"logits/rejected": -2.
|
422 |
-
"logps/chosen": -
|
423 |
-
"logps/rejected": -
|
424 |
-
"loss":
|
425 |
-
"rewards/accuracies": 0.
|
426 |
-
"rewards/chosen": -
|
427 |
-
"rewards/margins":
|
428 |
-
"rewards/rejected": -
|
429 |
"step": 270
|
430 |
},
|
431 |
{
|
432 |
-
"epoch": 0.
|
433 |
-
"grad_norm":
|
434 |
"learning_rate": 1.1398695954469597e-07,
|
435 |
-
"logits/chosen": -2.
|
436 |
-
"logits/rejected": -2.
|
437 |
-
"logps/chosen": -
|
438 |
-
"logps/rejected": -
|
439 |
-
"loss": 0.
|
440 |
-
"rewards/accuracies": 0.
|
441 |
-
"rewards/chosen":
|
442 |
-
"rewards/margins":
|
443 |
-
"rewards/rejected": -
|
444 |
"step": 280
|
445 |
},
|
446 |
{
|
447 |
-
"epoch": 0.
|
448 |
-
"grad_norm":
|
449 |
"learning_rate": 9.902380922818425e-08,
|
450 |
-
"logits/chosen": -2.
|
451 |
-
"logits/rejected": -2.
|
452 |
-
"logps/chosen": -
|
453 |
-
"logps/rejected": -
|
454 |
-
"loss": 0.
|
455 |
-
"rewards/accuracies": 0.
|
456 |
-
"rewards/chosen":
|
457 |
-
"rewards/margins":
|
458 |
-
"rewards/rejected": -
|
459 |
"step": 290
|
460 |
},
|
461 |
{
|
462 |
-
"epoch": 0.
|
463 |
-
"grad_norm":
|
464 |
"learning_rate": 8.478428028080398e-08,
|
465 |
-
"logits/chosen": -2.
|
466 |
-
"logits/rejected": -2.
|
467 |
-
"logps/chosen": -
|
468 |
-
"logps/rejected": -
|
469 |
-
"loss": 0.
|
470 |
-
"rewards/accuracies": 0.
|
471 |
-
"rewards/chosen": -
|
472 |
-
"rewards/margins":
|
473 |
-
"rewards/rejected": -
|
474 |
"step": 300
|
475 |
},
|
476 |
{
|
477 |
-
"epoch": 0.
|
478 |
-
"grad_norm":
|
479 |
"learning_rate": 7.137041683151202e-08,
|
480 |
-
"logits/chosen": -2.
|
481 |
-
"logits/rejected": -2.
|
482 |
-
"logps/chosen": -
|
483 |
-
"logps/rejected": -
|
484 |
-
"loss": 0.
|
485 |
-
"rewards/accuracies": 0.
|
486 |
-
"rewards/chosen": -
|
487 |
-
"rewards/margins":
|
488 |
-
"rewards/rejected": -5.
|
489 |
"step": 310
|
490 |
},
|
491 |
{
|
492 |
-
"epoch": 0.
|
493 |
-
"grad_norm":
|
494 |
"learning_rate": 5.8878346077822135e-08,
|
495 |
-
"logits/chosen": -2.
|
496 |
-
"logits/rejected": -2.
|
497 |
-
"logps/chosen": -
|
498 |
-
"logps/rejected": -
|
499 |
-
"loss": 0.
|
500 |
-
"rewards/accuracies": 0.
|
501 |
-
"rewards/chosen":
|
502 |
-
"rewards/margins":
|
503 |
-
"rewards/rejected": -
|
504 |
"step": 320
|
505 |
},
|
506 |
{
|
507 |
-
"epoch": 0.
|
508 |
-
"grad_norm":
|
509 |
"learning_rate": 4.73975894135696e-08,
|
510 |
-
"logits/chosen": -2.
|
511 |
-
"logits/rejected": -2.
|
512 |
-
"logps/chosen": -
|
513 |
-
"logps/rejected": -
|
514 |
-
"loss": 0.
|
515 |
-
"rewards/accuracies": 0.
|
516 |
-
"rewards/chosen": -
|
517 |
-
"rewards/margins":
|
518 |
-
"rewards/rejected": -4.
|
519 |
"step": 330
|
520 |
},
|
521 |
{
|
522 |
-
"epoch": 0.
|
523 |
-
"grad_norm":
|
524 |
"learning_rate": 3.701042089556483e-08,
|
525 |
-
"logits/chosen": -2.
|
526 |
-
"logits/rejected": -2.
|
527 |
-
"logps/chosen": -
|
528 |
-
"logps/rejected": -
|
529 |
-
"loss": 0.
|
530 |
-
"rewards/accuracies": 0.
|
531 |
-
"rewards/chosen": -
|
532 |
-
"rewards/margins":
|
533 |
-
"rewards/rejected": -
|
534 |
"step": 340
|
535 |
},
|
536 |
{
|
537 |
-
"epoch": 0.
|
538 |
-
"grad_norm":
|
539 |
"learning_rate": 2.779127764652889e-08,
|
540 |
-
"logits/chosen": -2.
|
541 |
-
"logits/rejected": -2.
|
542 |
-
"logps/chosen": -
|
543 |
-
"logps/rejected": -
|
544 |
-
"loss": 0.
|
545 |
-
"rewards/accuracies": 0.
|
546 |
-
"rewards/chosen": -
|
547 |
-
"rewards/margins":
|
548 |
-
"rewards/rejected": -4.
|
549 |
"step": 350
|
550 |
},
|
551 |
{
|
552 |
-
"epoch": 0.
|
553 |
-
"grad_norm":
|
554 |
"learning_rate": 1.9806226419516193e-08,
|
555 |
-
"logits/chosen": -2.
|
556 |
-
"logits/rejected": -2.
|
557 |
-
"logps/chosen": -
|
558 |
-
"logps/rejected": -
|
559 |
-
"loss": 0.
|
560 |
-
"rewards/accuracies": 0.
|
561 |
-
"rewards/chosen": -
|
562 |
-
"rewards/margins":
|
563 |
-
"rewards/rejected": -
|
564 |
"step": 360
|
565 |
},
|
566 |
{
|
567 |
-
"epoch": 0.
|
568 |
-
"grad_norm":
|
569 |
"learning_rate": 1.3112490146559552e-08,
|
570 |
-
"logits/chosen": -2.
|
571 |
-
"logits/rejected": -2.
|
572 |
-
"logps/chosen": -
|
573 |
-
"logps/rejected": -
|
574 |
-
"loss": 0.
|
575 |
-
"rewards/accuracies": 0.
|
576 |
-
"rewards/chosen": -
|
577 |
-
"rewards/margins":
|
578 |
-
"rewards/rejected": -4.
|
579 |
"step": 370
|
580 |
},
|
581 |
{
|
582 |
-
"epoch": 0.
|
583 |
-
"grad_norm":
|
584 |
"learning_rate": 7.758037864413247e-09,
|
585 |
-
"logits/chosen": -2.
|
586 |
-
"logits/rejected": -2.
|
587 |
-
"logps/chosen": -
|
588 |
-
"logps/rejected": -
|
589 |
-
"loss": 0.
|
590 |
-
"rewards/accuracies": 0.
|
591 |
-
"rewards/chosen": -
|
592 |
-
"rewards/margins":
|
593 |
-
"rewards/rejected": -
|
594 |
"step": 380
|
595 |
},
|
596 |
{
|
597 |
-
"epoch": 0.
|
598 |
-
"grad_norm":
|
599 |
"learning_rate": 3.78124095609087e-09,
|
600 |
-
"logits/chosen": -2.
|
601 |
-
"logits/rejected": -2.
|
602 |
-
"logps/chosen": -
|
603 |
-
"logps/rejected": -
|
604 |
-
"loss": 0.
|
605 |
-
"rewards/accuracies": 0.
|
606 |
-
"rewards/chosen":
|
607 |
-
"rewards/margins":
|
608 |
-
"rewards/rejected": -
|
609 |
"step": 390
|
610 |
},
|
611 |
{
|
612 |
-
"epoch": 0.
|
613 |
-
"grad_norm":
|
614 |
"learning_rate": 1.2105981716597603e-09,
|
615 |
-
"logits/chosen": -2.
|
616 |
-
"logits/rejected": -2.
|
617 |
-
"logps/chosen": -
|
618 |
-
"logps/rejected": -
|
619 |
-
"loss": 0.
|
620 |
-
"rewards/accuracies": 0.
|
621 |
-
"rewards/chosen": -
|
622 |
-
"rewards/margins":
|
623 |
-
"rewards/rejected": -
|
624 |
"step": 400
|
625 |
},
|
626 |
{
|
627 |
-
"epoch": 0.
|
628 |
-
"grad_norm":
|
629 |
"learning_rate": 6.453139886395398e-11,
|
630 |
-
"logits/chosen": -2.
|
631 |
-
"logits/rejected": -2.
|
632 |
-
"logps/chosen": -
|
633 |
-
"logps/rejected": -
|
634 |
-
"loss": 0.
|
635 |
-
"rewards/accuracies": 0.
|
636 |
-
"rewards/chosen": -
|
637 |
-
"rewards/margins":
|
638 |
-
"rewards/rejected": -4.
|
639 |
"step": 410
|
640 |
},
|
641 |
{
|
642 |
-
"epoch":
|
643 |
"step": 413,
|
644 |
"total_flos": 0.0,
|
645 |
-
"train_loss": 0.
|
646 |
-
"train_runtime":
|
647 |
-
"train_samples_per_second":
|
648 |
-
"train_steps_per_second": 0.
|
649 |
}
|
650 |
],
|
651 |
"logging_steps": 10,
|
@@ -653,6 +653,18 @@
|
|
653 |
"num_input_tokens_seen": 0,
|
654 |
"num_train_epochs": 1,
|
655 |
"save_steps": 100,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
656 |
"total_flos": 0.0,
|
657 |
"train_batch_size": 4,
|
658 |
"trial_name": null,
|
|
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
+
"epoch": 0.0024183796856106408,
|
13 |
+
"grad_norm": 1308.2534984366307,
|
14 |
"learning_rate": 9.523809523809522e-09,
|
15 |
"logits/chosen": -2.7005977630615234,
|
16 |
"logits/rejected": -2.6288318634033203,
|
|
|
24 |
"step": 1
|
25 |
},
|
26 |
{
|
27 |
+
"epoch": 0.02418379685610641,
|
28 |
+
"grad_norm": 1018.5726644948129,
|
29 |
"learning_rate": 9.523809523809523e-08,
|
30 |
+
"logits/chosen": -2.762432098388672,
|
31 |
+
"logits/rejected": -2.697216033935547,
|
32 |
+
"logps/chosen": -0.83743816614151,
|
33 |
+
"logps/rejected": -0.8184080123901367,
|
34 |
+
"loss": 0.7126,
|
35 |
+
"rewards/accuracies": 0.4652777910232544,
|
36 |
+
"rewards/chosen": 0.018050068989396095,
|
37 |
+
"rewards/margins": 0.10968472808599472,
|
38 |
+
"rewards/rejected": -0.09163466095924377,
|
39 |
"step": 10
|
40 |
},
|
41 |
{
|
42 |
+
"epoch": 0.04836759371221282,
|
43 |
+
"grad_norm": 1180.2159938574362,
|
44 |
"learning_rate": 1.9047619047619045e-07,
|
45 |
+
"logits/chosen": -2.6905813217163086,
|
46 |
+
"logits/rejected": -2.6509311199188232,
|
47 |
+
"logps/chosen": -0.9959298968315125,
|
48 |
+
"logps/rejected": -1.0417280197143555,
|
49 |
+
"loss": 0.6914,
|
50 |
+
"rewards/accuracies": 0.550000011920929,
|
51 |
+
"rewards/chosen": -0.41868314146995544,
|
52 |
+
"rewards/margins": 0.11889855563640594,
|
53 |
+
"rewards/rejected": -0.5375816822052002,
|
54 |
"step": 20
|
55 |
},
|
56 |
{
|
57 |
+
"epoch": 0.07255139056831923,
|
58 |
+
"grad_norm": 2875.4669192965785,
|
59 |
"learning_rate": 2.857142857142857e-07,
|
60 |
+
"logits/chosen": -2.7360830307006836,
|
61 |
+
"logits/rejected": -2.68190336227417,
|
62 |
+
"logps/chosen": -0.9741474986076355,
|
63 |
+
"logps/rejected": -0.9867057800292969,
|
64 |
+
"loss": 0.7598,
|
65 |
+
"rewards/accuracies": 0.6937500238418579,
|
66 |
+
"rewards/chosen": -0.7634207606315613,
|
67 |
+
"rewards/margins": 0.5467337965965271,
|
68 |
+
"rewards/rejected": -1.3101545572280884,
|
69 |
"step": 30
|
70 |
},
|
71 |
{
|
72 |
+
"epoch": 0.09673518742442563,
|
73 |
+
"grad_norm": 2637.428065513591,
|
74 |
"learning_rate": 3.809523809523809e-07,
|
75 |
+
"logits/chosen": -2.686359405517578,
|
76 |
+
"logits/rejected": -2.641474485397339,
|
77 |
+
"logps/chosen": -0.9992626309394836,
|
78 |
+
"logps/rejected": -0.9355602264404297,
|
79 |
+
"loss": 0.9431,
|
80 |
+
"rewards/accuracies": 0.762499988079071,
|
81 |
+
"rewards/chosen": -1.0757026672363281,
|
82 |
+
"rewards/margins": 1.4628146886825562,
|
83 |
+
"rewards/rejected": -2.538517475128174,
|
84 |
"step": 40
|
85 |
},
|
86 |
{
|
87 |
+
"epoch": 0.12091898428053205,
|
88 |
+
"grad_norm": 975.4162480488375,
|
89 |
"learning_rate": 3.995412608484087e-07,
|
90 |
+
"logits/chosen": -2.759364366531372,
|
91 |
+
"logits/rejected": -2.7036304473876953,
|
92 |
+
"logps/chosen": -0.9589918255805969,
|
93 |
+
"logps/rejected": -0.9145007133483887,
|
94 |
+
"loss": 0.9551,
|
95 |
+
"rewards/accuracies": 0.793749988079071,
|
96 |
+
"rewards/chosen": 0.7774861454963684,
|
97 |
+
"rewards/margins": 1.538771629333496,
|
98 |
+
"rewards/rejected": -0.7612855434417725,
|
99 |
"step": 50
|
100 |
},
|
101 |
{
|
102 |
+
"epoch": 0.14510278113663846,
|
103 |
+
"grad_norm": 876.8403861103245,
|
104 |
"learning_rate": 3.976812391485896e-07,
|
105 |
+
"logits/chosen": -2.762204647064209,
|
106 |
+
"logits/rejected": -2.693788528442383,
|
107 |
+
"logps/chosen": -0.9397481083869934,
|
108 |
+
"logps/rejected": -0.9560295939445496,
|
109 |
+
"loss": 1.0176,
|
110 |
+
"rewards/accuracies": 0.8125,
|
111 |
+
"rewards/chosen": -2.3399124145507812,
|
112 |
+
"rewards/margins": 2.62353777885437,
|
113 |
+
"rewards/rejected": -4.9634504318237305,
|
114 |
"step": 60
|
115 |
},
|
116 |
{
|
117 |
+
"epoch": 0.16928657799274485,
|
118 |
+
"grad_norm": 851.542671730339,
|
119 |
"learning_rate": 3.9440458281608213e-07,
|
120 |
+
"logits/chosen": -2.753873109817505,
|
121 |
+
"logits/rejected": -2.727172613143921,
|
122 |
+
"logps/chosen": -0.9090434312820435,
|
123 |
+
"logps/rejected": -0.8764857053756714,
|
124 |
+
"loss": 0.9609,
|
125 |
+
"rewards/accuracies": 0.8187500238418579,
|
126 |
+
"rewards/chosen": 1.77558171749115,
|
127 |
+
"rewards/margins": 3.0382332801818848,
|
128 |
+
"rewards/rejected": -1.2626516819000244,
|
129 |
"step": 70
|
130 |
},
|
131 |
{
|
132 |
+
"epoch": 0.19347037484885127,
|
133 |
+
"grad_norm": 1072.4102961713095,
|
134 |
"learning_rate": 3.897347732134074e-07,
|
135 |
+
"logits/chosen": -2.6825203895568848,
|
136 |
+
"logits/rejected": -2.6251988410949707,
|
137 |
+
"logps/chosen": -0.8969934582710266,
|
138 |
+
"logps/rejected": -1.0139261484146118,
|
139 |
+
"loss": 1.0729,
|
140 |
+
"rewards/accuracies": 0.824999988079071,
|
141 |
+
"rewards/chosen": -0.25338679552078247,
|
142 |
+
"rewards/margins": 2.662139415740967,
|
143 |
+
"rewards/rejected": -2.9155266284942627,
|
144 |
"step": 80
|
145 |
},
|
146 |
{
|
147 |
+
"epoch": 0.21765417170495768,
|
148 |
+
"grad_norm": 1045.2736360081997,
|
149 |
"learning_rate": 3.8370527539794614e-07,
|
150 |
+
"logits/chosen": -2.6782150268554688,
|
151 |
+
"logits/rejected": -2.6257784366607666,
|
152 |
+
"logps/chosen": -1.0229610204696655,
|
153 |
+
"logps/rejected": -1.0610243082046509,
|
154 |
+
"loss": 1.0661,
|
155 |
+
"rewards/accuracies": 0.737500011920929,
|
156 |
+
"rewards/chosen": -0.9386919736862183,
|
157 |
+
"rewards/margins": 3.1748452186584473,
|
158 |
+
"rewards/rejected": -4.113537311553955,
|
159 |
"step": 90
|
160 |
},
|
161 |
{
|
162 |
+
"epoch": 0.2418379685610641,
|
163 |
+
"grad_norm": 717.8046840728214,
|
164 |
"learning_rate": 3.763592983027255e-07,
|
165 |
+
"logits/chosen": -2.7070720195770264,
|
166 |
+
"logits/rejected": -2.657769203186035,
|
167 |
+
"logps/chosen": -0.8991826772689819,
|
168 |
+
"logps/rejected": -0.9345502853393555,
|
169 |
+
"loss": 1.1057,
|
170 |
+
"rewards/accuracies": 0.8062499761581421,
|
171 |
+
"rewards/chosen": -0.7811681032180786,
|
172 |
+
"rewards/margins": 2.7421443462371826,
|
173 |
+
"rewards/rejected": -3.52331280708313,
|
174 |
"step": 100
|
175 |
},
|
176 |
{
|
177 |
+
"epoch": 0.2660217654171705,
|
178 |
+
"grad_norm": 573.0671194598642,
|
179 |
"learning_rate": 3.6774948509008527e-07,
|
180 |
+
"logits/chosen": -2.715163469314575,
|
181 |
+
"logits/rejected": -2.6663975715637207,
|
182 |
+
"logps/chosen": -0.9820922613143921,
|
183 |
+
"logps/rejected": -0.9612447023391724,
|
184 |
+
"loss": 0.9831,
|
185 |
+
"rewards/accuracies": 0.8187500238418579,
|
186 |
+
"rewards/chosen": -0.3871256709098816,
|
187 |
+
"rewards/margins": 2.8536558151245117,
|
188 |
+
"rewards/rejected": -3.240781784057617,
|
189 |
"step": 110
|
190 |
},
|
191 |
{
|
192 |
+
"epoch": 0.29020556227327693,
|
193 |
+
"grad_norm": 813.0340299259154,
|
194 |
"learning_rate": 3.579375358972288e-07,
|
195 |
+
"logits/chosen": -2.667712688446045,
|
196 |
+
"logits/rejected": -2.6149513721466064,
|
197 |
+
"logps/chosen": -0.9040060043334961,
|
198 |
+
"logps/rejected": -1.0196150541305542,
|
199 |
+
"loss": 0.8569,
|
200 |
+
"rewards/accuracies": 0.8125,
|
201 |
+
"rewards/chosen": -2.3184046745300293,
|
202 |
+
"rewards/margins": 3.717514753341675,
|
203 |
+
"rewards/rejected": -6.035919189453125,
|
204 |
"step": 120
|
205 |
},
|
206 |
{
|
207 |
+
"epoch": 0.3143893591293833,
|
208 |
+
"grad_norm": 803.5906659383959,
|
209 |
"learning_rate": 3.4699376567716156e-07,
|
210 |
+
"logits/chosen": -2.7040085792541504,
|
211 |
+
"logits/rejected": -2.662682056427002,
|
212 |
+
"logps/chosen": -0.8633000254631042,
|
213 |
+
"logps/rejected": -0.8938501477241516,
|
214 |
+
"loss": 1.0665,
|
215 |
+
"rewards/accuracies": 0.8062499761581421,
|
216 |
+
"rewards/chosen": 0.6341629028320312,
|
217 |
+
"rewards/margins": 4.641495704650879,
|
218 |
+
"rewards/rejected": -4.007332801818848,
|
219 |
"step": 130
|
220 |
},
|
221 |
{
|
222 |
+
"epoch": 0.3385731559854897,
|
223 |
+
"grad_norm": 894.1357604383927,
|
224 |
"learning_rate": 3.349966003036421e-07,
|
225 |
+
"logits/chosen": -2.678401470184326,
|
226 |
+
"logits/rejected": -2.6329421997070312,
|
227 |
+
"logps/chosen": -0.9393678903579712,
|
228 |
+
"logps/rejected": -0.9559372663497925,
|
229 |
+
"loss": 0.9368,
|
230 |
+
"rewards/accuracies": 0.793749988079071,
|
231 |
+
"rewards/chosen": -1.3545489311218262,
|
232 |
+
"rewards/margins": 2.7926135063171387,
|
233 |
+
"rewards/rejected": -4.147162437438965,
|
234 |
"step": 140
|
235 |
},
|
236 |
{
|
237 |
+
"epoch": 0.36275695284159615,
|
238 |
+
"grad_norm": 1218.2734701044658,
|
239 |
"learning_rate": 3.220320145511884e-07,
|
240 |
+
"logits/chosen": -2.691074848175049,
|
241 |
+
"logits/rejected": -2.6277267932891846,
|
242 |
+
"logps/chosen": -0.9556415677070618,
|
243 |
+
"logps/rejected": -1.0070703029632568,
|
244 |
+
"loss": 1.0655,
|
245 |
+
"rewards/accuracies": 0.8187500238418579,
|
246 |
+
"rewards/chosen": 0.038965918123722076,
|
247 |
+
"rewards/margins": 3.301687717437744,
|
248 |
+
"rewards/rejected": -3.262721538543701,
|
249 |
"step": 150
|
250 |
},
|
251 |
{
|
252 |
+
"epoch": 0.38694074969770254,
|
253 |
+
"grad_norm": 933.333059653217,
|
254 |
"learning_rate": 3.0819291597771795e-07,
|
255 |
+
"logits/chosen": -2.6850905418395996,
|
256 |
+
"logits/rejected": -2.641737461090088,
|
257 |
+
"logps/chosen": -0.9150887727737427,
|
258 |
+
"logps/rejected": -0.9567450284957886,
|
259 |
+
"loss": 1.1619,
|
260 |
+
"rewards/accuracies": 0.824999988079071,
|
261 |
+
"rewards/chosen": -0.4916467070579529,
|
262 |
+
"rewards/margins": 3.2452170848846436,
|
263 |
+
"rewards/rejected": -3.7368640899658203,
|
264 |
"step": 160
|
265 |
},
|
266 |
{
|
267 |
+
"epoch": 0.4111245465538089,
|
268 |
+
"grad_norm": 1195.530681047137,
|
269 |
"learning_rate": 2.9357847912507786e-07,
|
270 |
+
"logits/chosen": -2.664726734161377,
|
271 |
+
"logits/rejected": -2.5912623405456543,
|
272 |
+
"logps/chosen": -0.908007800579071,
|
273 |
+
"logps/rejected": -0.9083712697029114,
|
274 |
+
"loss": 1.0815,
|
275 |
+
"rewards/accuracies": 0.831250011920929,
|
276 |
+
"rewards/chosen": -1.6012508869171143,
|
277 |
+
"rewards/margins": 3.351865768432617,
|
278 |
+
"rewards/rejected": -4.9531168937683105,
|
279 |
"step": 170
|
280 |
},
|
281 |
{
|
282 |
+
"epoch": 0.43530834340991537,
|
283 |
+
"grad_norm": 1688.331550337793,
|
284 |
"learning_rate": 2.7829343480875617e-07,
|
285 |
+
"logits/chosen": -2.662369966506958,
|
286 |
+
"logits/rejected": -2.5877506732940674,
|
287 |
+
"logps/chosen": -0.9355185627937317,
|
288 |
+
"logps/rejected": -0.9674497842788696,
|
289 |
+
"loss": 1.0193,
|
290 |
+
"rewards/accuracies": 0.84375,
|
291 |
+
"rewards/chosen": 0.7115722894668579,
|
292 |
+
"rewards/margins": 3.03961181640625,
|
293 |
+
"rewards/rejected": -2.3280396461486816,
|
294 |
"step": 180
|
295 |
},
|
296 |
{
|
297 |
+
"epoch": 0.45949214026602175,
|
298 |
+
"grad_norm": 1025.3772351835823,
|
299 |
"learning_rate": 2.624473195899052e-07,
|
300 |
+
"logits/chosen": -2.7287068367004395,
|
301 |
+
"logits/rejected": -2.696254253387451,
|
302 |
+
"logps/chosen": -0.9740760922431946,
|
303 |
+
"logps/rejected": -1.0570310354232788,
|
304 |
+
"loss": 1.0517,
|
305 |
+
"rewards/accuracies": 0.768750011920929,
|
306 |
+
"rewards/chosen": -1.7886015176773071,
|
307 |
+
"rewards/margins": 3.46142578125,
|
308 |
+
"rewards/rejected": -5.250028133392334,
|
309 |
"step": 190
|
310 |
},
|
311 |
{
|
312 |
+
"epoch": 0.4836759371221282,
|
313 |
+
"grad_norm": 781.90565238033,
|
314 |
"learning_rate": 2.4615369080815547e-07,
|
315 |
+
"logits/chosen": -2.6937649250030518,
|
316 |
+
"logits/rejected": -2.651794195175171,
|
317 |
+
"logps/chosen": -0.8561639785766602,
|
318 |
+
"logps/rejected": -0.9404775500297546,
|
319 |
+
"loss": 1.0482,
|
320 |
+
"rewards/accuracies": 0.831250011920929,
|
321 |
+
"rewards/chosen": 0.10315074771642685,
|
322 |
+
"rewards/margins": 2.2021610736846924,
|
323 |
+
"rewards/rejected": -2.0990102291107178,
|
324 |
"step": 200
|
325 |
},
|
326 |
{
|
327 |
+
"epoch": 0.5078597339782346,
|
328 |
+
"grad_norm": 718.5486788904052,
|
329 |
"learning_rate": 2.2952931280049625e-07,
|
330 |
+
"logits/chosen": -2.7320737838745117,
|
331 |
+
"logits/rejected": -2.666625499725342,
|
332 |
+
"logps/chosen": -1.0195786952972412,
|
333 |
+
"logps/rejected": -0.985085129737854,
|
334 |
+
"loss": 1.2414,
|
335 |
+
"rewards/accuracies": 0.8062499761581421,
|
336 |
+
"rewards/chosen": 0.24961701035499573,
|
337 |
+
"rewards/margins": 4.099853038787842,
|
338 |
+
"rewards/rejected": -3.850236415863037,
|
339 |
"step": 210
|
340 |
},
|
341 |
{
|
342 |
+
"epoch": 0.532043530834341,
|
343 |
+
"grad_norm": 896.7965591252768,
|
344 |
"learning_rate": 2.1269332013798747e-07,
|
345 |
+
"logits/chosen": -2.7412266731262207,
|
346 |
+
"logits/rejected": -2.719102144241333,
|
347 |
+
"logps/chosen": -0.887231171131134,
|
348 |
+
"logps/rejected": -0.8848485946655273,
|
349 |
+
"loss": 1.1268,
|
350 |
+
"rewards/accuracies": 0.7875000238418579,
|
351 |
+
"rewards/chosen": 0.5452920198440552,
|
352 |
+
"rewards/margins": 3.039072036743164,
|
353 |
+
"rewards/rejected": -2.4937801361083984,
|
354 |
"step": 220
|
355 |
},
|
356 |
{
|
357 |
+
"epoch": 0.5562273276904474,
|
358 |
+
"grad_norm": 606.0417693874598,
|
359 |
"learning_rate": 1.9576636387676436e-07,
|
360 |
+
"logits/chosen": -2.687635898590088,
|
361 |
+
"logits/rejected": -2.6457619667053223,
|
362 |
+
"logps/chosen": -0.9592903852462769,
|
363 |
+
"logps/rejected": -0.967265784740448,
|
364 |
+
"loss": 0.8874,
|
365 |
+
"rewards/accuracies": 0.75,
|
366 |
+
"rewards/chosen": -1.5358507633209229,
|
367 |
+
"rewards/margins": 3.3348402976989746,
|
368 |
+
"rewards/rejected": -4.870690822601318,
|
369 |
"step": 230
|
370 |
},
|
371 |
{
|
372 |
+
"epoch": 0.5804111245465539,
|
373 |
+
"grad_norm": 794.9088518264721,
|
374 |
"learning_rate": 1.7886974694151976e-07,
|
375 |
+
"logits/chosen": -2.701521158218384,
|
376 |
+
"logits/rejected": -2.6726996898651123,
|
377 |
+
"logps/chosen": -0.9976873397827148,
|
378 |
+
"logps/rejected": -1.01997709274292,
|
379 |
+
"loss": 1.0599,
|
380 |
+
"rewards/accuracies": 0.831250011920929,
|
381 |
+
"rewards/chosen": -0.8589717745780945,
|
382 |
+
"rewards/margins": 3.750645875930786,
|
383 |
+
"rewards/rejected": -4.609617710113525,
|
384 |
"step": 240
|
385 |
},
|
386 |
{
|
387 |
+
"epoch": 0.6045949214026602,
|
388 |
+
"grad_norm": 706.1920920085257,
|
389 |
"learning_rate": 1.6212455483752895e-07,
|
390 |
+
"logits/chosen": -2.7483222484588623,
|
391 |
+
"logits/rejected": -2.66450834274292,
|
392 |
+
"logps/chosen": -0.8861182928085327,
|
393 |
+
"logps/rejected": -0.9307917356491089,
|
394 |
+
"loss": 1.0161,
|
395 |
+
"rewards/accuracies": 0.8125,
|
396 |
+
"rewards/chosen": 1.0754988193511963,
|
397 |
+
"rewards/margins": 3.2043755054473877,
|
398 |
+
"rewards/rejected": -2.1288766860961914,
|
399 |
"step": 250
|
400 |
},
|
401 |
{
|
402 |
+
"epoch": 0.6287787182587666,
|
403 |
+
"grad_norm": 1160.850163008235,
|
404 |
"learning_rate": 1.4565078792075733e-07,
|
405 |
+
"logits/chosen": -2.699370861053467,
|
406 |
+
"logits/rejected": -2.6291871070861816,
|
407 |
+
"logps/chosen": -1.0056778192520142,
|
408 |
+
"logps/rejected": -1.0176211595535278,
|
409 |
+
"loss": 1.0258,
|
410 |
+
"rewards/accuracies": 0.8374999761581421,
|
411 |
+
"rewards/chosen": 0.27300015091896057,
|
412 |
+
"rewards/margins": 4.846710681915283,
|
413 |
+
"rewards/rejected": -4.5737104415893555,
|
414 |
"step": 260
|
415 |
},
|
416 |
{
|
417 |
+
"epoch": 0.652962515114873,
|
418 |
+
"grad_norm": 882.9319448662282,
|
419 |
"learning_rate": 1.295665014444281e-07,
|
420 |
+
"logits/chosen": -2.726219654083252,
|
421 |
+
"logits/rejected": -2.6420578956604004,
|
422 |
+
"logps/chosen": -0.9637600183486938,
|
423 |
+
"logps/rejected": -0.975050151348114,
|
424 |
+
"loss": 1.1212,
|
425 |
+
"rewards/accuracies": 0.862500011920929,
|
426 |
+
"rewards/chosen": -1.4779157638549805,
|
427 |
+
"rewards/margins": 3.851034641265869,
|
428 |
+
"rewards/rejected": -5.328949928283691,
|
429 |
"step": 270
|
430 |
},
|
431 |
{
|
432 |
+
"epoch": 0.6771463119709794,
|
433 |
+
"grad_norm": 914.3545207404874,
|
434 |
"learning_rate": 1.1398695954469597e-07,
|
435 |
+
"logits/chosen": -2.673750162124634,
|
436 |
+
"logits/rejected": -2.611456871032715,
|
437 |
+
"logps/chosen": -0.9046699404716492,
|
438 |
+
"logps/rejected": -0.9061228036880493,
|
439 |
+
"loss": 0.9744,
|
440 |
+
"rewards/accuracies": 0.8125,
|
441 |
+
"rewards/chosen": 0.9529012441635132,
|
442 |
+
"rewards/margins": 2.8092963695526123,
|
443 |
+
"rewards/rejected": -1.8563950061798096,
|
444 |
"step": 280
|
445 |
},
|
446 |
{
|
447 |
+
"epoch": 0.7013301088270859,
|
448 |
+
"grad_norm": 1001.1455181062619,
|
449 |
"learning_rate": 9.902380922818425e-08,
|
450 |
+
"logits/chosen": -2.7217631340026855,
|
451 |
+
"logits/rejected": -2.676079273223877,
|
452 |
+
"logps/chosen": -0.986601710319519,
|
453 |
+
"logps/rejected": -0.992447555065155,
|
454 |
+
"loss": 0.7804,
|
455 |
+
"rewards/accuracies": 0.824999988079071,
|
456 |
+
"rewards/chosen": 1.5409294366836548,
|
457 |
+
"rewards/margins": 3.72483491897583,
|
458 |
+
"rewards/rejected": -2.1839053630828857,
|
459 |
"step": 290
|
460 |
},
|
461 |
{
|
462 |
+
"epoch": 0.7255139056831923,
|
463 |
+
"grad_norm": 929.8360933885888,
|
464 |
"learning_rate": 8.478428028080398e-08,
|
465 |
+
"logits/chosen": -2.7178378105163574,
|
466 |
+
"logits/rejected": -2.6601595878601074,
|
467 |
+
"logps/chosen": -0.9026691317558289,
|
468 |
+
"logps/rejected": -0.9635857343673706,
|
469 |
+
"loss": 0.963,
|
470 |
+
"rewards/accuracies": 0.8062499761581421,
|
471 |
+
"rewards/chosen": -0.20143906772136688,
|
472 |
+
"rewards/margins": 3.291618824005127,
|
473 |
+
"rewards/rejected": -3.4930579662323,
|
474 |
"step": 300
|
475 |
},
|
476 |
{
|
477 |
+
"epoch": 0.7496977025392987,
|
478 |
+
"grad_norm": 918.556437892916,
|
479 |
"learning_rate": 7.137041683151202e-08,
|
480 |
+
"logits/chosen": -2.7046775817871094,
|
481 |
+
"logits/rejected": -2.6347708702087402,
|
482 |
+
"logps/chosen": -1.0870015621185303,
|
483 |
+
"logps/rejected": -1.049997091293335,
|
484 |
+
"loss": 0.7071,
|
485 |
+
"rewards/accuracies": 0.793749988079071,
|
486 |
+
"rewards/chosen": -1.3678367137908936,
|
487 |
+
"rewards/margins": 3.927046298980713,
|
488 |
+
"rewards/rejected": -5.294882774353027,
|
489 |
"step": 310
|
490 |
},
|
491 |
{
|
492 |
+
"epoch": 0.7738814993954051,
|
493 |
+
"grad_norm": 1444.2173436767062,
|
494 |
"learning_rate": 5.8878346077822135e-08,
|
495 |
+
"logits/chosen": -2.7104134559631348,
|
496 |
+
"logits/rejected": -2.6237189769744873,
|
497 |
+
"logps/chosen": -0.8994497060775757,
|
498 |
+
"logps/rejected": -0.9579949378967285,
|
499 |
+
"loss": 0.7542,
|
500 |
+
"rewards/accuracies": 0.90625,
|
501 |
+
"rewards/chosen": 0.5055674910545349,
|
502 |
+
"rewards/margins": 4.5611371994018555,
|
503 |
+
"rewards/rejected": -4.055570602416992,
|
504 |
"step": 320
|
505 |
},
|
506 |
{
|
507 |
+
"epoch": 0.7980652962515115,
|
508 |
+
"grad_norm": 758.7885560560964,
|
509 |
"learning_rate": 4.73975894135696e-08,
|
510 |
+
"logits/chosen": -2.6587905883789062,
|
511 |
+
"logits/rejected": -2.584770917892456,
|
512 |
+
"logps/chosen": -0.9313557744026184,
|
513 |
+
"logps/rejected": -0.9829689860343933,
|
514 |
+
"loss": 0.7798,
|
515 |
+
"rewards/accuracies": 0.862500011920929,
|
516 |
+
"rewards/chosen": -0.45623573660850525,
|
517 |
+
"rewards/margins": 4.346068859100342,
|
518 |
+
"rewards/rejected": -4.802304744720459,
|
519 |
"step": 330
|
520 |
},
|
521 |
{
|
522 |
+
"epoch": 0.8222490931076178,
|
523 |
+
"grad_norm": 818.3776763189691,
|
524 |
"learning_rate": 3.701042089556483e-08,
|
525 |
+
"logits/chosen": -2.7390825748443604,
|
526 |
+
"logits/rejected": -2.6632418632507324,
|
527 |
+
"logps/chosen": -0.9014676809310913,
|
528 |
+
"logps/rejected": -0.9521511197090149,
|
529 |
+
"loss": 0.8929,
|
530 |
+
"rewards/accuracies": 0.7749999761581421,
|
531 |
+
"rewards/chosen": -1.1090164184570312,
|
532 |
+
"rewards/margins": 2.5331737995147705,
|
533 |
+
"rewards/rejected": -3.642190456390381,
|
534 |
"step": 340
|
535 |
},
|
536 |
{
|
537 |
+
"epoch": 0.8464328899637243,
|
538 |
+
"grad_norm": 633.4468460505269,
|
539 |
"learning_rate": 2.779127764652889e-08,
|
540 |
+
"logits/chosen": -2.671653985977173,
|
541 |
+
"logits/rejected": -2.6091012954711914,
|
542 |
+
"logps/chosen": -0.9844328761100769,
|
543 |
+
"logps/rejected": -0.9899514317512512,
|
544 |
+
"loss": 0.8681,
|
545 |
+
"rewards/accuracies": 0.8125,
|
546 |
+
"rewards/chosen": -1.1147739887237549,
|
547 |
+
"rewards/margins": 3.46244478225708,
|
548 |
+
"rewards/rejected": -4.577218532562256,
|
549 |
"step": 350
|
550 |
},
|
551 |
{
|
552 |
+
"epoch": 0.8706166868198307,
|
553 |
+
"grad_norm": 790.3321831454036,
|
554 |
"learning_rate": 1.9806226419516193e-08,
|
555 |
+
"logits/chosen": -2.682281494140625,
|
556 |
+
"logits/rejected": -2.6280040740966797,
|
557 |
+
"logps/chosen": -0.9713428616523743,
|
558 |
+
"logps/rejected": -1.028427004814148,
|
559 |
+
"loss": 0.8939,
|
560 |
+
"rewards/accuracies": 0.7875000238418579,
|
561 |
+
"rewards/chosen": -0.3930521309375763,
|
562 |
+
"rewards/margins": 3.496274948120117,
|
563 |
+
"rewards/rejected": -3.889326810836792,
|
564 |
"step": 360
|
565 |
},
|
566 |
{
|
567 |
+
"epoch": 0.8948004836759371,
|
568 |
+
"grad_norm": 758.0022509460326,
|
569 |
"learning_rate": 1.3112490146559552e-08,
|
570 |
+
"logits/chosen": -2.7291171550750732,
|
571 |
+
"logits/rejected": -2.6654868125915527,
|
572 |
+
"logps/chosen": -0.8997148275375366,
|
573 |
+
"logps/rejected": -0.9173600077629089,
|
574 |
+
"loss": 0.6798,
|
575 |
+
"rewards/accuracies": 0.8812500238418579,
|
576 |
+
"rewards/chosen": -0.3203060030937195,
|
577 |
+
"rewards/margins": 3.888981580734253,
|
578 |
+
"rewards/rejected": -4.209287166595459,
|
579 |
"step": 370
|
580 |
},
|
581 |
{
|
582 |
+
"epoch": 0.9189842805320435,
|
583 |
+
"grad_norm": 580.6775247669598,
|
584 |
"learning_rate": 7.758037864413247e-09,
|
585 |
+
"logits/chosen": -2.6995849609375,
|
586 |
+
"logits/rejected": -2.6686558723449707,
|
587 |
+
"logps/chosen": -0.906518280506134,
|
588 |
+
"logps/rejected": -0.9883183240890503,
|
589 |
+
"loss": 0.7423,
|
590 |
+
"rewards/accuracies": 0.8687499761581421,
|
591 |
+
"rewards/chosen": -0.12804336845874786,
|
592 |
+
"rewards/margins": 3.583407163619995,
|
593 |
+
"rewards/rejected": -3.7114500999450684,
|
594 |
"step": 380
|
595 |
},
|
596 |
{
|
597 |
+
"epoch": 0.9431680773881499,
|
598 |
+
"grad_norm": 740.0766659575787,
|
599 |
"learning_rate": 3.78124095609087e-09,
|
600 |
+
"logits/chosen": -2.677032947540283,
|
601 |
+
"logits/rejected": -2.63029408454895,
|
602 |
+
"logps/chosen": -0.9278643727302551,
|
603 |
+
"logps/rejected": -1.0108085870742798,
|
604 |
+
"loss": 0.7435,
|
605 |
+
"rewards/accuracies": 0.84375,
|
606 |
+
"rewards/chosen": 0.13243435323238373,
|
607 |
+
"rewards/margins": 3.7296395301818848,
|
608 |
+
"rewards/rejected": -3.597205400466919,
|
609 |
"step": 390
|
610 |
},
|
611 |
{
|
612 |
+
"epoch": 0.9673518742442564,
|
613 |
+
"grad_norm": 664.330995388725,
|
614 |
"learning_rate": 1.2105981716597603e-09,
|
615 |
+
"logits/chosen": -2.711874485015869,
|
616 |
+
"logits/rejected": -2.614854574203491,
|
617 |
+
"logps/chosen": -0.9732038378715515,
|
618 |
+
"logps/rejected": -0.9536064863204956,
|
619 |
+
"loss": 0.955,
|
620 |
+
"rewards/accuracies": 0.831250011920929,
|
621 |
+
"rewards/chosen": -0.32332465052604675,
|
622 |
+
"rewards/margins": 3.985724925994873,
|
623 |
+
"rewards/rejected": -4.309049129486084,
|
624 |
"step": 400
|
625 |
},
|
626 |
{
|
627 |
+
"epoch": 0.9915356711003628,
|
628 |
+
"grad_norm": 1022.9589946634649,
|
629 |
"learning_rate": 6.453139886395398e-11,
|
630 |
+
"logits/chosen": -2.710538864135742,
|
631 |
+
"logits/rejected": -2.6645543575286865,
|
632 |
+
"logps/chosen": -0.9343970417976379,
|
633 |
+
"logps/rejected": -0.9787089228630066,
|
634 |
+
"loss": 0.8541,
|
635 |
+
"rewards/accuracies": 0.831250011920929,
|
636 |
+
"rewards/chosen": -0.5227594375610352,
|
637 |
+
"rewards/margins": 4.429939270019531,
|
638 |
+
"rewards/rejected": -4.952698707580566,
|
639 |
"step": 410
|
640 |
},
|
641 |
{
|
642 |
+
"epoch": 0.9987908101571947,
|
643 |
"step": 413,
|
644 |
"total_flos": 0.0,
|
645 |
+
"train_loss": 0.9438655665076674,
|
646 |
+
"train_runtime": 7020.7263,
|
647 |
+
"train_samples_per_second": 7.538,
|
648 |
+
"train_steps_per_second": 0.059
|
649 |
}
|
650 |
],
|
651 |
"logging_steps": 10,
|
|
|
653 |
"num_input_tokens_seen": 0,
|
654 |
"num_train_epochs": 1,
|
655 |
"save_steps": 100,
|
656 |
+
"stateful_callbacks": {
|
657 |
+
"TrainerControl": {
|
658 |
+
"args": {
|
659 |
+
"should_epoch_stop": false,
|
660 |
+
"should_evaluate": false,
|
661 |
+
"should_log": false,
|
662 |
+
"should_save": true,
|
663 |
+
"should_training_stop": false
|
664 |
+
},
|
665 |
+
"attributes": {}
|
666 |
+
}
|
667 |
+
},
|
668 |
"total_flos": 0.0,
|
669 |
"train_batch_size": 4,
|
670 |
"trial_name": null,
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c3a645c7a17ab4c7ed2624bb6f63a7530266096f7cd34d78eb56492568bf376c
|
3 |
+
size 6456
|