Model save
Browse files- README.md +2 -2
- all_results.json +5 -5
- 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/Jun14_02-05-29_n136-100-194/events.out.tfevents.1718302064.n136-100-194.643461.0 +2 -2
- train_results.json +5 -5
- trainer_state.json +240 -195
- training_args.bin +1 -1
README.md
CHANGED
@@ -32,7 +32,7 @@ More information needed
|
|
32 |
### Training hyperparameters
|
33 |
|
34 |
The following hyperparameters were used during training:
|
35 |
-
- learning_rate:
|
36 |
- train_batch_size: 8
|
37 |
- eval_batch_size: 8
|
38 |
- seed: 42
|
@@ -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 0.15.2
|
|
|
32 |
### Training hyperparameters
|
33 |
|
34 |
The following hyperparameters were used during training:
|
35 |
+
- learning_rate: 3e-07
|
36 |
- train_batch_size: 8
|
37 |
- eval_batch_size: 8
|
38 |
- seed: 42
|
|
|
52 |
|
53 |
### Framework versions
|
54 |
|
55 |
+
- Transformers 4.39.3
|
56 |
- Pytorch 2.1.2+cu118
|
57 |
- Datasets 2.16.1
|
58 |
- Tokenizers 0.15.2
|
all_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
-
"train_loss":
|
4 |
-
"train_runtime":
|
5 |
-
"train_samples":
|
6 |
-
"train_samples_per_second":
|
7 |
-
"train_steps_per_second": 0.
|
8 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"train_loss": 1.3861158726707337,
|
4 |
+
"train_runtime": 5367.071,
|
5 |
+
"train_samples": 48530,
|
6 |
+
"train_samples_per_second": 9.042,
|
7 |
+
"train_steps_per_second": 0.035
|
8 |
}
|
generation_config.json
CHANGED
@@ -2,5 +2,5 @@
|
|
2 |
"_from_model_config": true,
|
3 |
"bos_token_id": 1,
|
4 |
"eos_token_id": 32000,
|
5 |
-
"transformers_version": "4.
|
6 |
}
|
|
|
2 |
"_from_model_config": true,
|
3 |
"bos_token_id": 1,
|
4 |
"eos_token_id": 32000,
|
5 |
+
"transformers_version": "4.39.3"
|
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 4943178720
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8b368d145612a34ed994ad48082d9b257146213f25c1ebd309358d2894fc8166
|
3 |
size 4943178720
|
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 sha256:9b56652344c4b8a4db7f423c746179279fd145c51c280f7981952f11f7b406de
|
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 4540532728
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:76a444457bdcf58fff4f2801fa8105c617087ad5220718bb0eb8c51ce71f1022
|
3 |
size 4540532728
|
runs/Jun14_02-05-29_n136-100-194/events.out.tfevents.1718302064.n136-100-194.643461.0
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 sha256:51b1aeb1055493e96a48a15605d2fe768213e463fe5c236032745b917039c655
|
3 |
+
size 18173
|
train_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
-
"train_loss":
|
4 |
-
"train_runtime":
|
5 |
-
"train_samples":
|
6 |
-
"train_samples_per_second":
|
7 |
-
"train_steps_per_second": 0.
|
8 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"train_loss": 1.3861158726707337,
|
4 |
+
"train_runtime": 5367.071,
|
5 |
+
"train_samples": 48530,
|
6 |
+
"train_samples_per_second": 9.042,
|
7 |
+
"train_steps_per_second": 0.035
|
8 |
}
|
trainer_state.json
CHANGED
@@ -1,22 +1,22 @@
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
-
"epoch": 0.
|
5 |
"eval_steps": 500,
|
6 |
-
"global_step":
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
"epoch": 0.01,
|
13 |
-
"grad_norm":
|
14 |
-
"learning_rate":
|
15 |
-
"logits/chosen": -
|
16 |
-
"logits/rejected": -
|
17 |
-
"logps/chosen": -
|
18 |
-
"logps/rejected": -
|
19 |
-
"loss":
|
20 |
"rewards/accuracies": 0.0,
|
21 |
"rewards/chosen": 0.0,
|
22 |
"rewards/margins": 0.0,
|
@@ -24,242 +24,287 @@
|
|
24 |
"step": 1
|
25 |
},
|
26 |
{
|
27 |
-
"epoch": 0.
|
28 |
-
"grad_norm":
|
29 |
-
"learning_rate":
|
30 |
-
"logits/chosen": -
|
31 |
-
"logits/rejected": -
|
32 |
-
"logps/chosen": -
|
33 |
-
"logps/rejected": -
|
34 |
-
"loss":
|
35 |
-
"rewards/accuracies": 0.
|
36 |
-
"rewards/chosen": -0.
|
37 |
-
"rewards/margins": 0.
|
38 |
-
"rewards/rejected": -
|
39 |
"step": 10
|
40 |
},
|
41 |
{
|
42 |
-
"epoch": 0.
|
43 |
-
"grad_norm":
|
44 |
-
"learning_rate":
|
45 |
-
"logits/chosen": -
|
46 |
-
"logits/rejected": -
|
47 |
-
"logps/chosen": -
|
48 |
-
"logps/rejected": -
|
49 |
-
"loss":
|
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":
|
60 |
-
"logits/chosen": -
|
61 |
-
"logits/rejected": -
|
62 |
-
"logps/chosen": -
|
63 |
-
"logps/rejected": -
|
64 |
-
"loss":
|
65 |
-
"rewards/accuracies": 0.
|
66 |
-
"rewards/chosen": 0.
|
67 |
-
"rewards/margins": 0.
|
68 |
-
"rewards/rejected": -0.
|
69 |
"step": 30
|
70 |
},
|
71 |
{
|
72 |
-
"epoch": 0.
|
73 |
-
"grad_norm":
|
74 |
-
"learning_rate":
|
75 |
-
"logits/chosen": -
|
76 |
-
"logits/rejected": -
|
77 |
-
"logps/chosen": -
|
78 |
-
"logps/rejected": -
|
79 |
-
"loss":
|
80 |
-
"rewards/accuracies": 0.
|
81 |
-
"rewards/chosen": 0.
|
82 |
-
"rewards/margins": 0.
|
83 |
-
"rewards/rejected": -0.
|
84 |
"step": 40
|
85 |
},
|
86 |
{
|
87 |
-
"epoch": 0.
|
88 |
-
"grad_norm":
|
89 |
-
"learning_rate":
|
90 |
-
"logits/chosen": -
|
91 |
-
"logits/rejected": -
|
92 |
-
"logps/chosen": -
|
93 |
-
"logps/rejected": -
|
94 |
-
"loss":
|
95 |
-
"rewards/accuracies": 0.
|
96 |
-
"rewards/chosen": 0.
|
97 |
-
"rewards/margins":
|
98 |
-
"rewards/rejected": 0.
|
99 |
"step": 50
|
100 |
},
|
101 |
{
|
102 |
-
"epoch": 0.
|
103 |
-
"grad_norm":
|
104 |
-
"learning_rate":
|
105 |
-
"logits/chosen": -
|
106 |
-
"logits/rejected": -
|
107 |
-
"logps/chosen": -
|
108 |
-
"logps/rejected": -
|
109 |
-
"loss":
|
110 |
-
"rewards/accuracies": 0.
|
111 |
-
"rewards/chosen": 0.
|
112 |
-
"rewards/margins": 0.
|
113 |
-
"rewards/rejected":
|
114 |
"step": 60
|
115 |
},
|
116 |
{
|
117 |
-
"epoch": 0.
|
118 |
-
"grad_norm":
|
119 |
-
"learning_rate":
|
120 |
-
"logits/chosen": -
|
121 |
-
"logits/rejected": -
|
122 |
-
"logps/chosen": -
|
123 |
-
"logps/rejected": -
|
124 |
-
"loss":
|
125 |
-
"rewards/accuracies": 0.
|
126 |
-
"rewards/chosen":
|
127 |
-
"rewards/margins": 0.
|
128 |
-
"rewards/rejected":
|
129 |
"step": 70
|
130 |
},
|
131 |
{
|
132 |
-
"epoch": 0.
|
133 |
-
"grad_norm":
|
134 |
-
"learning_rate":
|
135 |
-
"logits/chosen": -
|
136 |
-
"logits/rejected": -
|
137 |
-
"logps/chosen": -
|
138 |
-
"logps/rejected": -
|
139 |
-
"loss":
|
140 |
-
"rewards/accuracies": 0.
|
141 |
-
"rewards/chosen":
|
142 |
-
"rewards/margins": 0.
|
143 |
-
"rewards/rejected":
|
144 |
"step": 80
|
145 |
},
|
146 |
{
|
147 |
-
"epoch": 0.
|
148 |
-
"grad_norm":
|
149 |
-
"learning_rate":
|
150 |
-
"logits/chosen": -4.
|
151 |
-
"logits/rejected": -4.
|
152 |
-
"logps/chosen": -
|
153 |
-
"logps/rejected": -
|
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":
|
165 |
-
"logits/chosen": -4.
|
166 |
-
"logits/rejected": -4.
|
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":
|
180 |
-
"logits/chosen": -4.
|
181 |
-
"logits/rejected": -4.
|
182 |
-
"logps/chosen": -
|
183 |
-
"logps/rejected": -
|
184 |
-
"loss":
|
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": 1.
|
195 |
-
"logits/chosen": -4.
|
196 |
-
"logits/rejected": -4.
|
197 |
-
"logps/chosen": -
|
198 |
-
"logps/rejected": -
|
199 |
-
"loss":
|
200 |
-
"rewards/accuracies": 0.
|
201 |
-
"rewards/chosen":
|
202 |
-
"rewards/margins":
|
203 |
-
"rewards/rejected":
|
204 |
"step": 120
|
205 |
},
|
206 |
{
|
207 |
-
"epoch": 0.
|
208 |
-
"grad_norm":
|
209 |
-
"learning_rate":
|
210 |
-
"logits/chosen": -4.
|
211 |
-
"logits/rejected": -4.
|
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":
|
225 |
-
"logits/chosen": -4.
|
226 |
-
"logits/rejected": -
|
227 |
-
"logps/chosen": -
|
228 |
-
"logps/rejected": -
|
229 |
-
"loss":
|
230 |
-
"rewards/accuracies": 0.
|
231 |
-
"rewards/chosen":
|
232 |
-
"rewards/margins":
|
233 |
-
"rewards/rejected":
|
234 |
"step": 140
|
235 |
},
|
236 |
{
|
237 |
-
"epoch":
|
238 |
-
"grad_norm":
|
239 |
-
"learning_rate":
|
240 |
-
"logits/chosen": -
|
241 |
-
"logits/rejected": -
|
242 |
-
"logps/chosen": -
|
243 |
-
"logps/rejected": -
|
244 |
-
"loss":
|
245 |
-
"rewards/accuracies": 0.
|
246 |
-
"rewards/chosen":
|
247 |
-
"rewards/margins":
|
248 |
-
"rewards/rejected":
|
249 |
"step": 150
|
250 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
251 |
{
|
252 |
"epoch": 1.0,
|
253 |
-
"step":
|
254 |
"total_flos": 0.0,
|
255 |
-
"train_loss":
|
256 |
-
"train_runtime":
|
257 |
-
"train_samples_per_second":
|
258 |
-
"train_steps_per_second": 0.
|
259 |
}
|
260 |
],
|
261 |
"logging_steps": 10,
|
262 |
-
"max_steps":
|
263 |
"num_input_tokens_seen": 0,
|
264 |
"num_train_epochs": 1,
|
265 |
"save_steps": 100,
|
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
+
"epoch": 0.9960474308300395,
|
5 |
"eval_steps": 500,
|
6 |
+
"global_step": 189,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
"epoch": 0.01,
|
13 |
+
"grad_norm": 41.432535799844544,
|
14 |
+
"learning_rate": 1.5789473684210525e-08,
|
15 |
+
"logits/chosen": -2.270329713821411,
|
16 |
+
"logits/rejected": -2.2495758533477783,
|
17 |
+
"logps/chosen": -1.0721262693405151,
|
18 |
+
"logps/rejected": -0.9967758059501648,
|
19 |
+
"loss": 1.8459,
|
20 |
"rewards/accuracies": 0.0,
|
21 |
"rewards/chosen": 0.0,
|
22 |
"rewards/margins": 0.0,
|
|
|
24 |
"step": 1
|
25 |
},
|
26 |
{
|
27 |
+
"epoch": 0.05,
|
28 |
+
"grad_norm": 38.0481408472616,
|
29 |
+
"learning_rate": 1.5789473684210525e-07,
|
30 |
+
"logits/chosen": -2.3005785942077637,
|
31 |
+
"logits/rejected": -2.236785888671875,
|
32 |
+
"logps/chosen": -1.0405129194259644,
|
33 |
+
"logps/rejected": -1.0305094718933105,
|
34 |
+
"loss": 1.8036,
|
35 |
+
"rewards/accuracies": 0.4618055522441864,
|
36 |
+
"rewards/chosen": -0.0005758580518886447,
|
37 |
+
"rewards/margins": -0.0005768582923337817,
|
38 |
+
"rewards/rejected": 1.000240445137024e-06,
|
39 |
"step": 10
|
40 |
},
|
41 |
{
|
42 |
+
"epoch": 0.11,
|
43 |
+
"grad_norm": 39.11632822940491,
|
44 |
+
"learning_rate": 2.9997438756870786e-07,
|
45 |
+
"logits/chosen": -2.323885917663574,
|
46 |
+
"logits/rejected": -2.296593189239502,
|
47 |
+
"logps/chosen": -1.053823709487915,
|
48 |
+
"logps/rejected": -1.0721189975738525,
|
49 |
+
"loss": 1.7888,
|
50 |
+
"rewards/accuracies": 0.690625011920929,
|
51 |
+
"rewards/chosen": -0.0061118630692362785,
|
52 |
+
"rewards/margins": 0.008750928565859795,
|
53 |
+
"rewards/rejected": -0.014862793497741222,
|
54 |
"step": 20
|
55 |
},
|
56 |
{
|
57 |
+
"epoch": 0.16,
|
58 |
+
"grad_norm": 34.61655659817795,
|
59 |
+
"learning_rate": 2.9691146514020486e-07,
|
60 |
+
"logits/chosen": -2.4537835121154785,
|
61 |
+
"logits/rejected": -2.406308889389038,
|
62 |
+
"logps/chosen": -1.035614013671875,
|
63 |
+
"logps/rejected": -1.090877890586853,
|
64 |
+
"loss": 1.7604,
|
65 |
+
"rewards/accuracies": 0.778124988079071,
|
66 |
+
"rewards/chosen": -0.04775990918278694,
|
67 |
+
"rewards/margins": 0.049577243626117706,
|
68 |
+
"rewards/rejected": -0.09733714908361435,
|
69 |
"step": 30
|
70 |
},
|
71 |
{
|
72 |
+
"epoch": 0.21,
|
73 |
+
"grad_norm": 39.56505728726439,
|
74 |
+
"learning_rate": 2.888456489672334e-07,
|
75 |
+
"logits/chosen": -2.672393321990967,
|
76 |
+
"logits/rejected": -2.6184394359588623,
|
77 |
+
"logps/chosen": -1.121576189994812,
|
78 |
+
"logps/rejected": -1.1622512340545654,
|
79 |
+
"loss": 1.7281,
|
80 |
+
"rewards/accuracies": 0.746874988079071,
|
81 |
+
"rewards/chosen": -0.1657789945602417,
|
82 |
+
"rewards/margins": 0.10218825191259384,
|
83 |
+
"rewards/rejected": -0.26796722412109375,
|
84 |
"step": 40
|
85 |
},
|
86 |
{
|
87 |
+
"epoch": 0.26,
|
88 |
+
"grad_norm": 41.60011182654463,
|
89 |
+
"learning_rate": 2.7605161074568387e-07,
|
90 |
+
"logits/chosen": -2.921407461166382,
|
91 |
+
"logits/rejected": -2.891125202178955,
|
92 |
+
"logps/chosen": -1.28288733959198,
|
93 |
+
"logps/rejected": -1.3418314456939697,
|
94 |
+
"loss": 1.7131,
|
95 |
+
"rewards/accuracies": 0.768750011920929,
|
96 |
+
"rewards/chosen": -0.4241597056388855,
|
97 |
+
"rewards/margins": 0.17930208146572113,
|
98 |
+
"rewards/rejected": -0.603461742401123,
|
99 |
"step": 50
|
100 |
},
|
101 |
{
|
102 |
+
"epoch": 0.32,
|
103 |
+
"grad_norm": 51.7670369092107,
|
104 |
+
"learning_rate": 2.5896503610243364e-07,
|
105 |
+
"logits/chosen": -3.184028148651123,
|
106 |
+
"logits/rejected": -3.152650833129883,
|
107 |
+
"logps/chosen": -1.467505693435669,
|
108 |
+
"logps/rejected": -1.6431095600128174,
|
109 |
+
"loss": 1.6172,
|
110 |
+
"rewards/accuracies": 0.7718750238418579,
|
111 |
+
"rewards/chosen": -0.7787442207336426,
|
112 |
+
"rewards/margins": 0.3277961313724518,
|
113 |
+
"rewards/rejected": -1.106540322303772,
|
114 |
"step": 60
|
115 |
},
|
116 |
{
|
117 |
+
"epoch": 0.37,
|
118 |
+
"grad_norm": 73.28833563129433,
|
119 |
+
"learning_rate": 2.3816778784387096e-07,
|
120 |
+
"logits/chosen": -3.472806930541992,
|
121 |
+
"logits/rejected": -3.4669137001037598,
|
122 |
+
"logps/chosen": -1.7310292720794678,
|
123 |
+
"logps/rejected": -1.9755589962005615,
|
124 |
+
"loss": 1.552,
|
125 |
+
"rewards/accuracies": 0.75,
|
126 |
+
"rewards/chosen": -1.4173996448516846,
|
127 |
+
"rewards/margins": 0.45889949798583984,
|
128 |
+
"rewards/rejected": -1.8762991428375244,
|
129 |
"step": 70
|
130 |
},
|
131 |
{
|
132 |
+
"epoch": 0.42,
|
133 |
+
"grad_norm": 95.75717630614442,
|
134 |
+
"learning_rate": 2.1436809131589132e-07,
|
135 |
+
"logits/chosen": -3.7732110023498535,
|
136 |
+
"logits/rejected": -3.794734477996826,
|
137 |
+
"logps/chosen": -2.3010740280151367,
|
138 |
+
"logps/rejected": -2.5936408042907715,
|
139 |
+
"loss": 1.4579,
|
140 |
+
"rewards/accuracies": 0.7093750238418579,
|
141 |
+
"rewards/chosen": -2.5355091094970703,
|
142 |
+
"rewards/margins": 0.5785154700279236,
|
143 |
+
"rewards/rejected": -3.1140246391296387,
|
144 |
"step": 80
|
145 |
},
|
146 |
{
|
147 |
+
"epoch": 0.47,
|
148 |
+
"grad_norm": 108.99770590059093,
|
149 |
+
"learning_rate": 1.8837641663916534e-07,
|
150 |
+
"logits/chosen": -4.047728061676025,
|
151 |
+
"logits/rejected": -4.063739776611328,
|
152 |
+
"logps/chosen": -2.8391122817993164,
|
153 |
+
"logps/rejected": -3.388620376586914,
|
154 |
+
"loss": 1.3563,
|
155 |
+
"rewards/accuracies": 0.734375,
|
156 |
+
"rewards/chosen": -3.563304901123047,
|
157 |
+
"rewards/margins": 1.0540244579315186,
|
158 |
+
"rewards/rejected": -4.617329120635986,
|
159 |
"step": 90
|
160 |
},
|
161 |
{
|
162 |
+
"epoch": 0.53,
|
163 |
+
"grad_norm": 91.4347139330851,
|
164 |
+
"learning_rate": 1.610778791212311e-07,
|
165 |
+
"logits/chosen": -4.239941596984863,
|
166 |
+
"logits/rejected": -4.29507303237915,
|
167 |
+
"logps/chosen": -2.653388738632202,
|
168 |
+
"logps/rejected": -3.2629570960998535,
|
169 |
+
"loss": 1.3233,
|
170 |
+
"rewards/accuracies": 0.778124988079071,
|
171 |
+
"rewards/chosen": -3.28294038772583,
|
172 |
+
"rewards/margins": 1.1777704954147339,
|
173 |
+
"rewards/rejected": -4.4607110023498535,
|
174 |
"step": 100
|
175 |
},
|
176 |
{
|
177 |
+
"epoch": 0.58,
|
178 |
+
"grad_norm": 131.69525214561466,
|
179 |
+
"learning_rate": 1.3340209771627488e-07,
|
180 |
+
"logits/chosen": -4.439741611480713,
|
181 |
+
"logits/rejected": -4.547017574310303,
|
182 |
+
"logps/chosen": -2.945267915725708,
|
183 |
+
"logps/rejected": -3.710526943206787,
|
184 |
+
"loss": 1.2097,
|
185 |
+
"rewards/accuracies": 0.778124988079071,
|
186 |
+
"rewards/chosen": -3.7780189514160156,
|
187 |
+
"rewards/margins": 1.5611803531646729,
|
188 |
+
"rewards/rejected": -5.339199542999268,
|
189 |
"step": 110
|
190 |
},
|
191 |
{
|
192 |
+
"epoch": 0.63,
|
193 |
+
"grad_norm": 119.65234559570642,
|
194 |
+
"learning_rate": 1.0629153796660131e-07,
|
195 |
+
"logits/chosen": -4.577895164489746,
|
196 |
+
"logits/rejected": -4.687996864318848,
|
197 |
+
"logps/chosen": -3.094693899154663,
|
198 |
+
"logps/rejected": -3.9398090839385986,
|
199 |
+
"loss": 1.1282,
|
200 |
+
"rewards/accuracies": 0.7749999761581421,
|
201 |
+
"rewards/chosen": -4.051518440246582,
|
202 |
+
"rewards/margins": 1.6234314441680908,
|
203 |
+
"rewards/rejected": -5.674950122833252,
|
204 |
"step": 120
|
205 |
},
|
206 |
{
|
207 |
+
"epoch": 0.69,
|
208 |
+
"grad_norm": 129.7843781353164,
|
209 |
+
"learning_rate": 8.066941746895304e-08,
|
210 |
+
"logits/chosen": -4.77476167678833,
|
211 |
+
"logits/rejected": -4.9267988204956055,
|
212 |
+
"logps/chosen": -3.1812126636505127,
|
213 |
+
"logps/rejected": -4.0755157470703125,
|
214 |
+
"loss": 1.1661,
|
215 |
+
"rewards/accuracies": 0.78125,
|
216 |
+
"rewards/chosen": -4.220904350280762,
|
217 |
+
"rewards/margins": 1.8181953430175781,
|
218 |
+
"rewards/rejected": -6.03909969329834,
|
219 |
"step": 130
|
220 |
},
|
221 |
{
|
222 |
+
"epoch": 0.74,
|
223 |
+
"grad_norm": 105.67370924537627,
|
224 |
+
"learning_rate": 5.7408266806531073e-08,
|
225 |
+
"logits/chosen": -4.906655311584473,
|
226 |
+
"logits/rejected": -5.071400165557861,
|
227 |
+
"logps/chosen": -3.282118320465088,
|
228 |
+
"logps/rejected": -4.2006425857543945,
|
229 |
+
"loss": 1.1872,
|
230 |
+
"rewards/accuracies": 0.778124988079071,
|
231 |
+
"rewards/chosen": -4.442627906799316,
|
232 |
+
"rewards/margins": 1.8388820886611938,
|
233 |
+
"rewards/rejected": -6.281510353088379,
|
234 |
"step": 140
|
235 |
},
|
236 |
{
|
237 |
+
"epoch": 0.79,
|
238 |
+
"grad_norm": 108.86520544080616,
|
239 |
+
"learning_rate": 3.730021656646899e-08,
|
240 |
+
"logits/chosen": -5.13236141204834,
|
241 |
+
"logits/rejected": -5.272187232971191,
|
242 |
+
"logps/chosen": -3.555389881134033,
|
243 |
+
"logps/rejected": -4.640769958496094,
|
244 |
+
"loss": 1.1411,
|
245 |
+
"rewards/accuracies": 0.7718750238418579,
|
246 |
+
"rewards/chosen": -5.0327253341674805,
|
247 |
+
"rewards/margins": 2.142836332321167,
|
248 |
+
"rewards/rejected": -7.175561428070068,
|
249 |
"step": 150
|
250 |
},
|
251 |
+
{
|
252 |
+
"epoch": 0.84,
|
253 |
+
"grad_norm": 104.57387509776791,
|
254 |
+
"learning_rate": 2.103002228274413e-08,
|
255 |
+
"logits/chosen": -4.998870372772217,
|
256 |
+
"logits/rejected": -5.1569671630859375,
|
257 |
+
"logps/chosen": -3.4073867797851562,
|
258 |
+
"logps/rejected": -4.47383975982666,
|
259 |
+
"loss": 1.1059,
|
260 |
+
"rewards/accuracies": 0.7718750238418579,
|
261 |
+
"rewards/chosen": -4.684414386749268,
|
262 |
+
"rewards/margins": 2.137641429901123,
|
263 |
+
"rewards/rejected": -6.822054862976074,
|
264 |
+
"step": 160
|
265 |
+
},
|
266 |
+
{
|
267 |
+
"epoch": 0.9,
|
268 |
+
"grad_norm": 109.01559488214998,
|
269 |
+
"learning_rate": 9.151745907741537e-09,
|
270 |
+
"logits/chosen": -5.156809329986572,
|
271 |
+
"logits/rejected": -5.336338520050049,
|
272 |
+
"logps/chosen": -3.476139545440674,
|
273 |
+
"logps/rejected": -4.648383617401123,
|
274 |
+
"loss": 1.113,
|
275 |
+
"rewards/accuracies": 0.8125,
|
276 |
+
"rewards/chosen": -4.8243327140808105,
|
277 |
+
"rewards/margins": 2.336613178253174,
|
278 |
+
"rewards/rejected": -7.160945892333984,
|
279 |
+
"step": 170
|
280 |
+
},
|
281 |
+
{
|
282 |
+
"epoch": 0.95,
|
283 |
+
"grad_norm": 149.6092511595762,
|
284 |
+
"learning_rate": 2.069887896989614e-09,
|
285 |
+
"logits/chosen": -5.088743209838867,
|
286 |
+
"logits/rejected": -5.213286399841309,
|
287 |
+
"logps/chosen": -3.5473315715789795,
|
288 |
+
"logps/rejected": -4.448221206665039,
|
289 |
+
"loss": 1.113,
|
290 |
+
"rewards/accuracies": 0.753125011920929,
|
291 |
+
"rewards/chosen": -5.046046733856201,
|
292 |
+
"rewards/margins": 1.7532546520233154,
|
293 |
+
"rewards/rejected": -6.799302101135254,
|
294 |
+
"step": 180
|
295 |
+
},
|
296 |
{
|
297 |
"epoch": 1.0,
|
298 |
+
"step": 189,
|
299 |
"total_flos": 0.0,
|
300 |
+
"train_loss": 1.3861158726707337,
|
301 |
+
"train_runtime": 5367.071,
|
302 |
+
"train_samples_per_second": 9.042,
|
303 |
+
"train_steps_per_second": 0.035
|
304 |
}
|
305 |
],
|
306 |
"logging_steps": 10,
|
307 |
+
"max_steps": 189,
|
308 |
"num_input_tokens_seen": 0,
|
309 |
"num_train_epochs": 1,
|
310 |
"save_steps": 100,
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 6264
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d677e274df19d2bbf6b42de6cc0260c9ddcb86589cf3edab1fca386d8bd8d657
|
3 |
size 6264
|