Upload folder using huggingface_hub
Browse files- adapter_model.bin +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- special_tokens_map.json +3 -4
- tokenizer.json +11 -2
- trainer_state.json +176 -68
- training_args.bin +1 -1
adapter_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 134264397
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b764cefe431eb39be2ea102916fa8d800b347c52d860823f8b22ca08bd55747d
|
| 3 |
size 134264397
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 268514437
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2a269b2a09b370d5ff9b834b7ad50bb4f5b1cc9a4b4f2914cef6c90826618422
|
| 3 |
size 268514437
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14575
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d3b60cfa0010d2dc843abda4bf592429e0146eda3f69771df5e72928c1bd85f1
|
| 3 |
size 14575
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a03bd55ec967989abd72e176bb55c4192fe56bafe1ccfdaea2dc4dd09d252069
|
| 3 |
size 627
|
special_tokens_map.json
CHANGED
|
@@ -1,10 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"additional_special_tokens": [
|
| 3 |
-
"<|im_end|>",
|
| 4 |
-
"<|im_start|>",
|
| 5 |
"<|end_of_turn|>",
|
| 6 |
-
"
|
| 7 |
-
"
|
|
|
|
| 8 |
],
|
| 9 |
"bos_token": "<|endoftext|>",
|
| 10 |
"eos_token": "<|endoftext|>",
|
|
|
|
| 1 |
{
|
| 2 |
"additional_special_tokens": [
|
|
|
|
|
|
|
| 3 |
"<|end_of_turn|>",
|
| 4 |
+
"<|start_of_turn|>",
|
| 5 |
+
"### Human:",
|
| 6 |
+
"### Assistant:"
|
| 7 |
],
|
| 8 |
"bos_token": "<|endoftext|>",
|
| 9 |
"eos_token": "<|endoftext|>",
|
tokenizer.json
CHANGED
|
@@ -289,7 +289,7 @@
|
|
| 289 |
},
|
| 290 |
{
|
| 291 |
"id": 50283,
|
| 292 |
-
"content": "
|
| 293 |
"single_word": false,
|
| 294 |
"lstrip": false,
|
| 295 |
"rstrip": false,
|
|
@@ -298,7 +298,16 @@
|
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"id": 50284,
|
| 301 |
-
"content": "
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 302 |
"single_word": false,
|
| 303 |
"lstrip": false,
|
| 304 |
"rstrip": false,
|
|
|
|
| 289 |
},
|
| 290 |
{
|
| 291 |
"id": 50283,
|
| 292 |
+
"content": "<|start_of_turn|>",
|
| 293 |
"single_word": false,
|
| 294 |
"lstrip": false,
|
| 295 |
"rstrip": false,
|
|
|
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"id": 50284,
|
| 301 |
+
"content": "### Human:",
|
| 302 |
+
"single_word": false,
|
| 303 |
+
"lstrip": false,
|
| 304 |
+
"rstrip": false,
|
| 305 |
+
"normalized": false,
|
| 306 |
+
"special": true
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"id": 50285,
|
| 310 |
+
"content": "### Assistant:",
|
| 311 |
"single_word": false,
|
| 312 |
"lstrip": false,
|
| 313 |
"rstrip": false,
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 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,
|
|
@@ -11,201 +11,309 @@
|
|
| 11 |
{
|
| 12 |
"epoch": 0.0,
|
| 13 |
"learning_rate": 2e-05,
|
| 14 |
-
"loss": 1.
|
| 15 |
"step": 2
|
| 16 |
},
|
| 17 |
{
|
| 18 |
-
"epoch": 0.
|
| 19 |
"learning_rate": 2e-05,
|
| 20 |
-
"loss": 1.
|
| 21 |
"step": 4
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch": 0.
|
| 25 |
"learning_rate": 2e-05,
|
| 26 |
-
"loss": 1.
|
| 27 |
"step": 6
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
"learning_rate": 2e-05,
|
| 32 |
-
"loss":
|
| 33 |
"step": 8
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch": 0.
|
| 37 |
"learning_rate": 2e-05,
|
| 38 |
-
"loss": 1.
|
| 39 |
"step": 10
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
"learning_rate": 2e-05,
|
| 44 |
-
"loss":
|
| 45 |
"step": 12
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
"learning_rate": 2e-05,
|
| 50 |
-
"loss":
|
| 51 |
"step": 14
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
"learning_rate": 2e-05,
|
| 56 |
-
"loss": 1.
|
| 57 |
"step": 16
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch": 0.
|
| 61 |
"learning_rate": 2e-05,
|
| 62 |
-
"loss": 1.
|
| 63 |
"step": 18
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"epoch": 0.
|
| 67 |
"learning_rate": 2e-05,
|
| 68 |
-
"loss":
|
| 69 |
"step": 20
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch": 0.
|
| 73 |
"learning_rate": 2e-05,
|
| 74 |
-
"loss":
|
| 75 |
"step": 22
|
| 76 |
},
|
| 77 |
{
|
| 78 |
-
"epoch": 0.
|
| 79 |
"learning_rate": 2e-05,
|
| 80 |
-
"loss":
|
| 81 |
"step": 24
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch": 0.
|
| 85 |
"learning_rate": 2e-05,
|
| 86 |
-
"loss":
|
| 87 |
"step": 26
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch": 0.
|
| 91 |
"learning_rate": 2e-05,
|
| 92 |
-
"loss":
|
| 93 |
"step": 28
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0.
|
| 97 |
"learning_rate": 2e-05,
|
| 98 |
-
"loss":
|
| 99 |
"step": 30
|
| 100 |
},
|
| 101 |
{
|
| 102 |
-
"epoch": 0.
|
| 103 |
"learning_rate": 2e-05,
|
| 104 |
-
"loss":
|
| 105 |
"step": 32
|
| 106 |
},
|
| 107 |
{
|
| 108 |
-
"epoch": 0.
|
| 109 |
"learning_rate": 2e-05,
|
| 110 |
-
"loss":
|
| 111 |
"step": 34
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch": 0.
|
| 115 |
"learning_rate": 2e-05,
|
| 116 |
-
"loss":
|
| 117 |
"step": 36
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch": 0.
|
| 121 |
"learning_rate": 2e-05,
|
| 122 |
-
"loss": 2.
|
| 123 |
"step": 38
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch": 0.
|
| 127 |
"learning_rate": 2e-05,
|
| 128 |
-
"loss": 2.
|
| 129 |
"step": 40
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch": 0.
|
| 133 |
"learning_rate": 2e-05,
|
| 134 |
-
"loss": 2.
|
| 135 |
"step": 42
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch": 0.
|
| 139 |
"learning_rate": 2e-05,
|
| 140 |
-
"loss":
|
| 141 |
"step": 44
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch": 0.
|
| 145 |
"learning_rate": 2e-05,
|
| 146 |
-
"loss":
|
| 147 |
"step": 46
|
| 148 |
},
|
| 149 |
{
|
| 150 |
-
"epoch": 0.
|
| 151 |
"learning_rate": 2e-05,
|
| 152 |
-
"loss":
|
| 153 |
"step": 48
|
| 154 |
},
|
| 155 |
{
|
| 156 |
-
"epoch": 0.
|
| 157 |
"learning_rate": 2e-05,
|
| 158 |
-
"loss":
|
| 159 |
"step": 50
|
| 160 |
},
|
| 161 |
{
|
| 162 |
-
"epoch": 0.
|
| 163 |
"learning_rate": 2e-05,
|
| 164 |
-
"loss": 1.
|
| 165 |
"step": 52
|
| 166 |
},
|
| 167 |
{
|
| 168 |
-
"epoch": 0.
|
| 169 |
"learning_rate": 2e-05,
|
| 170 |
-
"loss": 1.
|
| 171 |
"step": 54
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch": 0.
|
| 175 |
"learning_rate": 2e-05,
|
| 176 |
-
"loss": 1.
|
| 177 |
"step": 56
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch": 0.
|
| 181 |
"learning_rate": 2e-05,
|
| 182 |
-
"loss": 1.
|
| 183 |
"step": 58
|
| 184 |
},
|
| 185 |
{
|
| 186 |
-
"epoch": 0.
|
| 187 |
"learning_rate": 2e-05,
|
| 188 |
-
"loss": 1.
|
| 189 |
"step": 60
|
| 190 |
},
|
| 191 |
{
|
| 192 |
-
"epoch": 0.
|
| 193 |
"learning_rate": 2e-05,
|
| 194 |
-
"loss": 1.
|
| 195 |
"step": 62
|
| 196 |
},
|
| 197 |
{
|
| 198 |
-
"epoch": 0.
|
| 199 |
"learning_rate": 2e-05,
|
| 200 |
-
"loss": 1.
|
| 201 |
"step": 64
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 202 |
}
|
| 203 |
],
|
| 204 |
"logging_steps": 2,
|
| 205 |
-
"max_steps":
|
| 206 |
"num_train_epochs": 3,
|
| 207 |
-
"save_steps":
|
| 208 |
-
"total_flos":
|
| 209 |
"trial_name": null,
|
| 210 |
"trial_params": null
|
| 211 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.16250253910217347,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 100,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 11 |
{
|
| 12 |
"epoch": 0.0,
|
| 13 |
"learning_rate": 2e-05,
|
| 14 |
+
"loss": 1.7252,
|
| 15 |
"step": 2
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.01,
|
| 19 |
"learning_rate": 2e-05,
|
| 20 |
+
"loss": 1.8477,
|
| 21 |
"step": 4
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.01,
|
| 25 |
"learning_rate": 2e-05,
|
| 26 |
+
"loss": 1.6071,
|
| 27 |
"step": 6
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.01,
|
| 31 |
"learning_rate": 2e-05,
|
| 32 |
+
"loss": 2.0372,
|
| 33 |
"step": 8
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 0.02,
|
| 37 |
"learning_rate": 2e-05,
|
| 38 |
+
"loss": 1.9284,
|
| 39 |
"step": 10
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.02,
|
| 43 |
"learning_rate": 2e-05,
|
| 44 |
+
"loss": 2.0431,
|
| 45 |
"step": 12
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.02,
|
| 49 |
"learning_rate": 2e-05,
|
| 50 |
+
"loss": 2.1371,
|
| 51 |
"step": 14
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.03,
|
| 55 |
"learning_rate": 2e-05,
|
| 56 |
+
"loss": 1.9433,
|
| 57 |
"step": 16
|
| 58 |
},
|
| 59 |
{
|
| 60 |
+
"epoch": 0.03,
|
| 61 |
"learning_rate": 2e-05,
|
| 62 |
+
"loss": 1.987,
|
| 63 |
"step": 18
|
| 64 |
},
|
| 65 |
{
|
| 66 |
+
"epoch": 0.03,
|
| 67 |
"learning_rate": 2e-05,
|
| 68 |
+
"loss": 2.1397,
|
| 69 |
"step": 20
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 0.04,
|
| 73 |
"learning_rate": 2e-05,
|
| 74 |
+
"loss": 2.0042,
|
| 75 |
"step": 22
|
| 76 |
},
|
| 77 |
{
|
| 78 |
+
"epoch": 0.04,
|
| 79 |
"learning_rate": 2e-05,
|
| 80 |
+
"loss": 2.0193,
|
| 81 |
"step": 24
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 0.04,
|
| 85 |
"learning_rate": 2e-05,
|
| 86 |
+
"loss": 2.1728,
|
| 87 |
"step": 26
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 0.05,
|
| 91 |
"learning_rate": 2e-05,
|
| 92 |
+
"loss": 2.4132,
|
| 93 |
"step": 28
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.05,
|
| 97 |
"learning_rate": 2e-05,
|
| 98 |
+
"loss": 2.2364,
|
| 99 |
"step": 30
|
| 100 |
},
|
| 101 |
{
|
| 102 |
+
"epoch": 0.05,
|
| 103 |
"learning_rate": 2e-05,
|
| 104 |
+
"loss": 2.3455,
|
| 105 |
"step": 32
|
| 106 |
},
|
| 107 |
{
|
| 108 |
+
"epoch": 0.06,
|
| 109 |
"learning_rate": 2e-05,
|
| 110 |
+
"loss": 2.3668,
|
| 111 |
"step": 34
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 0.06,
|
| 115 |
"learning_rate": 2e-05,
|
| 116 |
+
"loss": 2.3953,
|
| 117 |
"step": 36
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 0.06,
|
| 121 |
"learning_rate": 2e-05,
|
| 122 |
+
"loss": 2.5691,
|
| 123 |
"step": 38
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 0.07,
|
| 127 |
"learning_rate": 2e-05,
|
| 128 |
+
"loss": 2.685,
|
| 129 |
"step": 40
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 0.07,
|
| 133 |
"learning_rate": 2e-05,
|
| 134 |
+
"loss": 2.7927,
|
| 135 |
"step": 42
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.07,
|
| 139 |
"learning_rate": 2e-05,
|
| 140 |
+
"loss": 2.9057,
|
| 141 |
"step": 44
|
| 142 |
},
|
| 143 |
{
|
| 144 |
+
"epoch": 0.07,
|
| 145 |
"learning_rate": 2e-05,
|
| 146 |
+
"loss": 3.2075,
|
| 147 |
"step": 46
|
| 148 |
},
|
| 149 |
{
|
| 150 |
+
"epoch": 0.08,
|
| 151 |
"learning_rate": 2e-05,
|
| 152 |
+
"loss": 3.5339,
|
| 153 |
"step": 48
|
| 154 |
},
|
| 155 |
{
|
| 156 |
+
"epoch": 0.08,
|
| 157 |
"learning_rate": 2e-05,
|
| 158 |
+
"loss": 4.2786,
|
| 159 |
"step": 50
|
| 160 |
},
|
| 161 |
{
|
| 162 |
+
"epoch": 0.08,
|
| 163 |
"learning_rate": 2e-05,
|
| 164 |
+
"loss": 1.7911,
|
| 165 |
"step": 52
|
| 166 |
},
|
| 167 |
{
|
| 168 |
+
"epoch": 0.09,
|
| 169 |
"learning_rate": 2e-05,
|
| 170 |
+
"loss": 1.6751,
|
| 171 |
"step": 54
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 0.09,
|
| 175 |
"learning_rate": 2e-05,
|
| 176 |
+
"loss": 1.7581,
|
| 177 |
"step": 56
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 0.09,
|
| 181 |
"learning_rate": 2e-05,
|
| 182 |
+
"loss": 1.783,
|
| 183 |
"step": 58
|
| 184 |
},
|
| 185 |
{
|
| 186 |
+
"epoch": 0.1,
|
| 187 |
"learning_rate": 2e-05,
|
| 188 |
+
"loss": 1.8982,
|
| 189 |
"step": 60
|
| 190 |
},
|
| 191 |
{
|
| 192 |
+
"epoch": 0.1,
|
| 193 |
"learning_rate": 2e-05,
|
| 194 |
+
"loss": 1.926,
|
| 195 |
"step": 62
|
| 196 |
},
|
| 197 |
{
|
| 198 |
+
"epoch": 0.1,
|
| 199 |
"learning_rate": 2e-05,
|
| 200 |
+
"loss": 1.8847,
|
| 201 |
"step": 64
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.11,
|
| 205 |
+
"learning_rate": 2e-05,
|
| 206 |
+
"loss": 2.0012,
|
| 207 |
+
"step": 66
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.11,
|
| 211 |
+
"learning_rate": 2e-05,
|
| 212 |
+
"loss": 2.0488,
|
| 213 |
+
"step": 68
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.11,
|
| 217 |
+
"learning_rate": 2e-05,
|
| 218 |
+
"loss": 2.042,
|
| 219 |
+
"step": 70
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.12,
|
| 223 |
+
"learning_rate": 2e-05,
|
| 224 |
+
"loss": 2.0934,
|
| 225 |
+
"step": 72
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.12,
|
| 229 |
+
"learning_rate": 2e-05,
|
| 230 |
+
"loss": 1.9756,
|
| 231 |
+
"step": 74
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.12,
|
| 235 |
+
"learning_rate": 2e-05,
|
| 236 |
+
"loss": 1.8503,
|
| 237 |
+
"step": 76
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.13,
|
| 241 |
+
"learning_rate": 2e-05,
|
| 242 |
+
"loss": 2.1736,
|
| 243 |
+
"step": 78
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.13,
|
| 247 |
+
"learning_rate": 2e-05,
|
| 248 |
+
"loss": 2.089,
|
| 249 |
+
"step": 80
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.13,
|
| 253 |
+
"learning_rate": 2e-05,
|
| 254 |
+
"loss": 2.3295,
|
| 255 |
+
"step": 82
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.14,
|
| 259 |
+
"learning_rate": 2e-05,
|
| 260 |
+
"loss": 2.2796,
|
| 261 |
+
"step": 84
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.14,
|
| 265 |
+
"learning_rate": 2e-05,
|
| 266 |
+
"loss": 2.2786,
|
| 267 |
+
"step": 86
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.14,
|
| 271 |
+
"learning_rate": 2e-05,
|
| 272 |
+
"loss": 2.2979,
|
| 273 |
+
"step": 88
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.15,
|
| 277 |
+
"learning_rate": 2e-05,
|
| 278 |
+
"loss": 2.5102,
|
| 279 |
+
"step": 90
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.15,
|
| 283 |
+
"learning_rate": 2e-05,
|
| 284 |
+
"loss": 2.6685,
|
| 285 |
+
"step": 92
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.15,
|
| 289 |
+
"learning_rate": 2e-05,
|
| 290 |
+
"loss": 2.7812,
|
| 291 |
+
"step": 94
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.16,
|
| 295 |
+
"learning_rate": 2e-05,
|
| 296 |
+
"loss": 2.943,
|
| 297 |
+
"step": 96
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.16,
|
| 301 |
+
"learning_rate": 2e-05,
|
| 302 |
+
"loss": 2.9432,
|
| 303 |
+
"step": 98
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.16,
|
| 307 |
+
"learning_rate": 2e-05,
|
| 308 |
+
"loss": 3.6691,
|
| 309 |
+
"step": 100
|
| 310 |
}
|
| 311 |
],
|
| 312 |
"logging_steps": 2,
|
| 313 |
+
"max_steps": 1845,
|
| 314 |
"num_train_epochs": 3,
|
| 315 |
+
"save_steps": 50,
|
| 316 |
+
"total_flos": 2179598319624192.0,
|
| 317 |
"trial_name": null,
|
| 318 |
"trial_params": null
|
| 319 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4027
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:174c8bffd086ebc6d1aa8c350cefa932511681b2eff2b9dd236c0f93f68017d2
|
| 3 |
size 4027
|