Upload 8 files
Browse files- config.json +54 -32
- model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +66 -1266
- training_args.bin +3 -0
config.json
CHANGED
|
@@ -1,20 +1,16 @@
|
|
| 1 |
{
|
| 2 |
"_name_or_path": "ayjays132/CustomGPT2Conversational",
|
| 3 |
-
"
|
| 4 |
"architectures": [
|
| 5 |
"GPT2LMHeadModel"
|
| 6 |
],
|
| 7 |
-
"
|
| 8 |
-
|
| 9 |
-
"eos_token_id": 50256,
|
| 10 |
-
"n_positions": 2048,
|
| 11 |
-
"padding_side": "left",
|
| 12 |
-
"truncation_side": "right"
|
| 13 |
-
},
|
| 14 |
"config": {
|
| 15 |
"activation_function": "gelu_new",
|
| 16 |
"attn_pdrop": 0.1,
|
| 17 |
"embd_pdrop": 0.1,
|
|
|
|
| 18 |
"initializer_range": 0.02,
|
| 19 |
"layer_norm_epsilon": 1e-05,
|
| 20 |
"n_ctx": 2048,
|
|
@@ -23,40 +19,66 @@
|
|
| 23 |
"n_layer": 36,
|
| 24 |
"n_positions": 2048,
|
| 25 |
"resid_pdrop": 0.1,
|
| 26 |
-
"vocab_size": 50257,
|
| 27 |
"scale_attn_weights": true,
|
| 28 |
-
"
|
| 29 |
-
"
|
| 30 |
-
},
|
| 31 |
-
"task_specific_params": {
|
| 32 |
-
"conversational": {
|
| 33 |
-
"max_length": 1024,
|
| 34 |
-
"min_length": 20,
|
| 35 |
-
"temperature": 0.7,
|
| 36 |
-
"top_k": 40,
|
| 37 |
-
"top_p": 0.95,
|
| 38 |
-
"num_beams": 5,
|
| 39 |
-
"no_repeat_ngram_size": 3,
|
| 40 |
-
"early_stopping": true,
|
| 41 |
-
"length_penalty": 2.0,
|
| 42 |
-
"do_sample": true,
|
| 43 |
-
"frequency_penalty": 0.5,
|
| 44 |
-
"presence_penalty": 0.5
|
| 45 |
-
}
|
| 46 |
},
|
| 47 |
-
"
|
| 48 |
-
"
|
| 49 |
-
"
|
| 50 |
-
"library_name": "transformers",
|
| 51 |
"language": "en",
|
|
|
|
|
|
|
| 52 |
"license": "apache-2.0",
|
| 53 |
"metrics": [
|
| 54 |
"perplexity",
|
| 55 |
"accuracy"
|
| 56 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 57 |
"pipeline_tag": "conversational",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 58 |
"tags": [
|
| 59 |
"conversational",
|
| 60 |
"state-of-the-art"
|
| 61 |
-
]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 62 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"_name_or_path": "ayjays132/CustomGPT2Conversational",
|
| 3 |
+
"activation_function": "gelu_new",
|
| 4 |
"architectures": [
|
| 5 |
"GPT2LMHeadModel"
|
| 6 |
],
|
| 7 |
+
"attn_pdrop": 0.1,
|
| 8 |
+
"bos_token_id": 50256,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
"config": {
|
| 10 |
"activation_function": "gelu_new",
|
| 11 |
"attn_pdrop": 0.1,
|
| 12 |
"embd_pdrop": 0.1,
|
| 13 |
+
"gradient_checkpointing": true,
|
| 14 |
"initializer_range": 0.02,
|
| 15 |
"layer_norm_epsilon": 1e-05,
|
| 16 |
"n_ctx": 2048,
|
|
|
|
| 19 |
"n_layer": 36,
|
| 20 |
"n_positions": 2048,
|
| 21 |
"resid_pdrop": 0.1,
|
|
|
|
| 22 |
"scale_attn_weights": true,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 50257
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 25 |
},
|
| 26 |
+
"embd_pdrop": 0.1,
|
| 27 |
+
"eos_token_id": 50256,
|
| 28 |
+
"initializer_range": 0.02,
|
|
|
|
| 29 |
"language": "en",
|
| 30 |
+
"layer_norm_epsilon": 1e-05,
|
| 31 |
+
"library_name": "transformers",
|
| 32 |
"license": "apache-2.0",
|
| 33 |
"metrics": [
|
| 34 |
"perplexity",
|
| 35 |
"accuracy"
|
| 36 |
],
|
| 37 |
+
"model_type": "gpt2",
|
| 38 |
+
"n_embd": 768,
|
| 39 |
+
"n_head": 12,
|
| 40 |
+
"n_inner": null,
|
| 41 |
+
"n_layer": 12,
|
| 42 |
+
"n_positions": 1024,
|
| 43 |
"pipeline_tag": "conversational",
|
| 44 |
+
"reorder_and_upcast_attn": false,
|
| 45 |
+
"resid_pdrop": 0.1,
|
| 46 |
+
"scale_attn_by_inverse_layer_idx": false,
|
| 47 |
+
"scale_attn_weights": true,
|
| 48 |
+
"summary_activation": null,
|
| 49 |
+
"summary_first_dropout": 0.1,
|
| 50 |
+
"summary_proj_to_labels": true,
|
| 51 |
+
"summary_type": "cls_index",
|
| 52 |
+
"summary_use_proj": true,
|
| 53 |
"tags": [
|
| 54 |
"conversational",
|
| 55 |
"state-of-the-art"
|
| 56 |
+
],
|
| 57 |
+
"task_specific_params": {
|
| 58 |
+
"conversational": {
|
| 59 |
+
"do_sample": true,
|
| 60 |
+
"early_stopping": true,
|
| 61 |
+
"frequency_penalty": 0.5,
|
| 62 |
+
"length_penalty": 2.0,
|
| 63 |
+
"max_length": 1024,
|
| 64 |
+
"min_length": 20,
|
| 65 |
+
"no_repeat_ngram_size": 3,
|
| 66 |
+
"num_beams": 5,
|
| 67 |
+
"presence_penalty": 0.5,
|
| 68 |
+
"temperature": 0.7,
|
| 69 |
+
"top_k": 40,
|
| 70 |
+
"top_p": 0.95
|
| 71 |
+
}
|
| 72 |
+
},
|
| 73 |
+
"tokenizer_config": {
|
| 74 |
+
"bos_token_id": 50256,
|
| 75 |
+
"eos_token_id": 50256,
|
| 76 |
+
"n_positions": 2048,
|
| 77 |
+
"padding_side": "left",
|
| 78 |
+
"truncation_side": "right"
|
| 79 |
+
},
|
| 80 |
+
"torch_dtype": "float32",
|
| 81 |
+
"transformers_version": "4.37.2",
|
| 82 |
+
"use_cache": true,
|
| 83 |
+
"vocab_size": 50257
|
| 84 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 497774208
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:839abdabb9e2db67117691de736daa0762f8cb9711b916429643a38f4010e757
|
| 3 |
size 497774208
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 995642298
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c69c7a2ea4b08b80736806da4a78af087235909bc337021f37381bad74760a90
|
| 3 |
size 995642298
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1dede6f4c6fec6e15aded13fe8fa4e3b188a548f0e43da285ecede4459eea5de
|
| 3 |
size 14244
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b4247e447234772e3730d02b6da20c344c682ce3e67dce3060fc52e6286ef180
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,1341 +1,141 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 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.
|
| 13 |
-
"learning_rate": 4.
|
| 14 |
-
"loss": 1.
|
| 15 |
"step": 500
|
| 16 |
},
|
| 17 |
{
|
| 18 |
-
"epoch": 0.
|
| 19 |
-
"learning_rate": 4.
|
| 20 |
-
"loss": 1.
|
| 21 |
"step": 1000
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch": 0.
|
| 25 |
-
"learning_rate": 4.
|
| 26 |
-
"loss": 1.
|
| 27 |
"step": 1500
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
-
"learning_rate": 4.
|
| 32 |
-
"loss": 1.
|
| 33 |
"step": 2000
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch": 0.
|
| 37 |
-
"learning_rate": 4.
|
| 38 |
-
"loss": 1.
|
| 39 |
"step": 2500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss": 1.
|
| 45 |
"step": 3000
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss": 1.
|
| 51 |
"step": 3500
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"learning_rate":
|
| 56 |
-
"loss": 1.
|
| 57 |
"step": 4000
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch": 0.
|
| 61 |
-
"learning_rate":
|
| 62 |
-
"loss": 1.
|
| 63 |
"step": 4500
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"epoch":
|
| 67 |
-
"learning_rate":
|
| 68 |
-
"loss": 1.
|
| 69 |
"step": 5000
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch":
|
| 73 |
-
"learning_rate":
|
| 74 |
-
"loss": 1.
|
| 75 |
"step": 5500
|
| 76 |
},
|
| 77 |
{
|
| 78 |
-
"epoch":
|
| 79 |
-
"learning_rate":
|
| 80 |
-
"loss": 1.
|
| 81 |
"step": 6000
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch":
|
| 85 |
-
"learning_rate":
|
| 86 |
-
"loss": 1.
|
| 87 |
"step": 6500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch":
|
| 91 |
-
"learning_rate":
|
| 92 |
-
"loss": 1.
|
| 93 |
"step": 7000
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch":
|
| 97 |
-
"learning_rate":
|
| 98 |
-
"loss": 1.
|
| 99 |
"step": 7500
|
| 100 |
},
|
| 101 |
{
|
| 102 |
-
"epoch":
|
| 103 |
-
"learning_rate":
|
| 104 |
-
"loss": 1.
|
| 105 |
"step": 8000
|
| 106 |
},
|
| 107 |
{
|
| 108 |
-
"epoch":
|
| 109 |
-
"learning_rate":
|
| 110 |
-
"loss": 1.
|
| 111 |
"step": 8500
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch":
|
| 115 |
-
"learning_rate":
|
| 116 |
-
"loss": 1.
|
| 117 |
"step": 9000
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch":
|
| 121 |
-
"learning_rate":
|
| 122 |
-
"loss": 1.
|
| 123 |
"step": 9500
|
| 124 |
},
|
| 125 |
-
{
|
| 126 |
-
"epoch": 1.08,
|
| 127 |
-
"learning_rate": 4.730777514538015e-05,
|
| 128 |
-
"loss": 1.7166,
|
| 129 |
-
"step": 10000
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 1.13,
|
| 133 |
-
"learning_rate": 4.7173163902649154e-05,
|
| 134 |
-
"loss": 1.7223,
|
| 135 |
-
"step": 10500
|
| 136 |
-
},
|
| 137 |
-
{
|
| 138 |
-
"epoch": 1.18,
|
| 139 |
-
"learning_rate": 4.7038552659918154e-05,
|
| 140 |
-
"loss": 1.7497,
|
| 141 |
-
"step": 11000
|
| 142 |
-
},
|
| 143 |
-
{
|
| 144 |
-
"epoch": 1.24,
|
| 145 |
-
"learning_rate": 4.690394141718716e-05,
|
| 146 |
-
"loss": 1.7416,
|
| 147 |
-
"step": 11500
|
| 148 |
-
},
|
| 149 |
-
{
|
| 150 |
-
"epoch": 1.29,
|
| 151 |
-
"learning_rate": 4.6769330174456175e-05,
|
| 152 |
-
"loss": 1.7603,
|
| 153 |
-
"step": 12000
|
| 154 |
-
},
|
| 155 |
-
{
|
| 156 |
-
"epoch": 1.35,
|
| 157 |
-
"learning_rate": 4.663471893172518e-05,
|
| 158 |
-
"loss": 1.7939,
|
| 159 |
-
"step": 12500
|
| 160 |
-
},
|
| 161 |
-
{
|
| 162 |
-
"epoch": 1.4,
|
| 163 |
-
"learning_rate": 4.650010768899419e-05,
|
| 164 |
-
"loss": 1.8245,
|
| 165 |
-
"step": 13000
|
| 166 |
-
},
|
| 167 |
-
{
|
| 168 |
-
"epoch": 1.45,
|
| 169 |
-
"learning_rate": 4.6365496446263195e-05,
|
| 170 |
-
"loss": 1.7997,
|
| 171 |
-
"step": 13500
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 1.51,
|
| 175 |
-
"learning_rate": 4.62308852035322e-05,
|
| 176 |
-
"loss": 1.8118,
|
| 177 |
-
"step": 14000
|
| 178 |
-
},
|
| 179 |
-
{
|
| 180 |
-
"epoch": 1.56,
|
| 181 |
-
"learning_rate": 4.609627396080121e-05,
|
| 182 |
-
"loss": 1.8217,
|
| 183 |
-
"step": 14500
|
| 184 |
-
},
|
| 185 |
-
{
|
| 186 |
-
"epoch": 1.62,
|
| 187 |
-
"learning_rate": 4.5961662718070216e-05,
|
| 188 |
-
"loss": 1.813,
|
| 189 |
-
"step": 15000
|
| 190 |
-
},
|
| 191 |
-
{
|
| 192 |
-
"epoch": 1.67,
|
| 193 |
-
"learning_rate": 4.582705147533922e-05,
|
| 194 |
-
"loss": 1.8473,
|
| 195 |
-
"step": 15500
|
| 196 |
-
},
|
| 197 |
-
{
|
| 198 |
-
"epoch": 1.72,
|
| 199 |
-
"learning_rate": 4.569244023260823e-05,
|
| 200 |
-
"loss": 1.8383,
|
| 201 |
-
"step": 16000
|
| 202 |
-
},
|
| 203 |
-
{
|
| 204 |
-
"epoch": 1.78,
|
| 205 |
-
"learning_rate": 4.555782898987724e-05,
|
| 206 |
-
"loss": 1.851,
|
| 207 |
-
"step": 16500
|
| 208 |
-
},
|
| 209 |
-
{
|
| 210 |
-
"epoch": 1.83,
|
| 211 |
-
"learning_rate": 4.5423217747146244e-05,
|
| 212 |
-
"loss": 1.869,
|
| 213 |
-
"step": 17000
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 1.88,
|
| 217 |
-
"learning_rate": 4.528860650441525e-05,
|
| 218 |
-
"loss": 1.8694,
|
| 219 |
-
"step": 17500
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 1.94,
|
| 223 |
-
"learning_rate": 4.515399526168426e-05,
|
| 224 |
-
"loss": 1.8716,
|
| 225 |
-
"step": 18000
|
| 226 |
-
},
|
| 227 |
-
{
|
| 228 |
-
"epoch": 1.99,
|
| 229 |
-
"learning_rate": 4.5019384018953264e-05,
|
| 230 |
-
"loss": 1.8382,
|
| 231 |
-
"step": 18500
|
| 232 |
-
},
|
| 233 |
-
{
|
| 234 |
-
"epoch": 2.05,
|
| 235 |
-
"learning_rate": 4.488477277622227e-05,
|
| 236 |
-
"loss": 1.6506,
|
| 237 |
-
"step": 19000
|
| 238 |
-
},
|
| 239 |
-
{
|
| 240 |
-
"epoch": 2.1,
|
| 241 |
-
"learning_rate": 4.475016153349128e-05,
|
| 242 |
-
"loss": 1.6231,
|
| 243 |
-
"step": 19500
|
| 244 |
-
},
|
| 245 |
{
|
| 246 |
"epoch": 2.15,
|
| 247 |
-
"learning_rate":
|
| 248 |
-
"loss":
|
| 249 |
-
"step":
|
| 250 |
-
},
|
| 251 |
-
{
|
| 252 |
-
"epoch": 2.21,
|
| 253 |
-
"learning_rate": 4.448093904802929e-05,
|
| 254 |
-
"loss": 1.6543,
|
| 255 |
-
"step": 20500
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 2.26,
|
| 259 |
-
"learning_rate": 4.43463278052983e-05,
|
| 260 |
-
"loss": 1.6816,
|
| 261 |
-
"step": 21000
|
| 262 |
-
},
|
| 263 |
-
{
|
| 264 |
-
"epoch": 2.32,
|
| 265 |
-
"learning_rate": 4.4211716562567306e-05,
|
| 266 |
-
"loss": 1.6582,
|
| 267 |
-
"step": 21500
|
| 268 |
-
},
|
| 269 |
-
{
|
| 270 |
-
"epoch": 2.37,
|
| 271 |
-
"learning_rate": 4.407710531983631e-05,
|
| 272 |
-
"loss": 1.6787,
|
| 273 |
-
"step": 22000
|
| 274 |
-
},
|
| 275 |
-
{
|
| 276 |
-
"epoch": 2.42,
|
| 277 |
-
"learning_rate": 4.394249407710532e-05,
|
| 278 |
-
"loss": 1.7136,
|
| 279 |
-
"step": 22500
|
| 280 |
-
},
|
| 281 |
-
{
|
| 282 |
-
"epoch": 2.48,
|
| 283 |
-
"learning_rate": 4.380788283437433e-05,
|
| 284 |
-
"loss": 1.7168,
|
| 285 |
-
"step": 23000
|
| 286 |
-
},
|
| 287 |
-
{
|
| 288 |
-
"epoch": 2.53,
|
| 289 |
-
"learning_rate": 4.367327159164334e-05,
|
| 290 |
-
"loss": 1.6984,
|
| 291 |
-
"step": 23500
|
| 292 |
-
},
|
| 293 |
-
{
|
| 294 |
-
"epoch": 2.58,
|
| 295 |
-
"learning_rate": 4.353866034891234e-05,
|
| 296 |
-
"loss": 1.7052,
|
| 297 |
-
"step": 24000
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 2.64,
|
| 301 |
-
"learning_rate": 4.340404910618135e-05,
|
| 302 |
-
"loss": 1.7213,
|
| 303 |
-
"step": 24500
|
| 304 |
-
},
|
| 305 |
-
{
|
| 306 |
-
"epoch": 2.69,
|
| 307 |
-
"learning_rate": 4.3269437863450354e-05,
|
| 308 |
-
"loss": 1.7305,
|
| 309 |
-
"step": 25000
|
| 310 |
-
},
|
| 311 |
-
{
|
| 312 |
-
"epoch": 2.75,
|
| 313 |
-
"learning_rate": 4.313482662071937e-05,
|
| 314 |
-
"loss": 1.7348,
|
| 315 |
-
"step": 25500
|
| 316 |
-
},
|
| 317 |
-
{
|
| 318 |
-
"epoch": 2.8,
|
| 319 |
-
"learning_rate": 4.3000215377988375e-05,
|
| 320 |
-
"loss": 1.7359,
|
| 321 |
-
"step": 26000
|
| 322 |
-
},
|
| 323 |
-
{
|
| 324 |
-
"epoch": 2.85,
|
| 325 |
-
"learning_rate": 4.286560413525738e-05,
|
| 326 |
-
"loss": 1.7278,
|
| 327 |
-
"step": 26500
|
| 328 |
-
},
|
| 329 |
-
{
|
| 330 |
-
"epoch": 2.91,
|
| 331 |
-
"learning_rate": 4.273099289252638e-05,
|
| 332 |
-
"loss": 1.7558,
|
| 333 |
-
"step": 27000
|
| 334 |
-
},
|
| 335 |
-
{
|
| 336 |
-
"epoch": 2.96,
|
| 337 |
-
"learning_rate": 4.259638164979539e-05,
|
| 338 |
-
"loss": 1.7545,
|
| 339 |
-
"step": 27500
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 3.02,
|
| 343 |
-
"learning_rate": 4.24617704070644e-05,
|
| 344 |
-
"loss": 1.6658,
|
| 345 |
-
"step": 28000
|
| 346 |
-
},
|
| 347 |
-
{
|
| 348 |
-
"epoch": 3.07,
|
| 349 |
-
"learning_rate": 4.232715916433341e-05,
|
| 350 |
-
"loss": 1.4971,
|
| 351 |
-
"step": 28500
|
| 352 |
-
},
|
| 353 |
-
{
|
| 354 |
-
"epoch": 3.12,
|
| 355 |
-
"learning_rate": 4.2192547921602416e-05,
|
| 356 |
-
"loss": 1.519,
|
| 357 |
-
"step": 29000
|
| 358 |
-
},
|
| 359 |
-
{
|
| 360 |
-
"epoch": 3.18,
|
| 361 |
-
"learning_rate": 4.205793667887142e-05,
|
| 362 |
-
"loss": 1.5378,
|
| 363 |
-
"step": 29500
|
| 364 |
-
},
|
| 365 |
-
{
|
| 366 |
-
"epoch": 3.23,
|
| 367 |
-
"learning_rate": 4.192332543614042e-05,
|
| 368 |
-
"loss": 1.556,
|
| 369 |
-
"step": 30000
|
| 370 |
-
},
|
| 371 |
-
{
|
| 372 |
-
"epoch": 3.28,
|
| 373 |
-
"learning_rate": 4.1788714193409437e-05,
|
| 374 |
-
"loss": 1.5828,
|
| 375 |
-
"step": 30500
|
| 376 |
-
},
|
| 377 |
-
{
|
| 378 |
-
"epoch": 3.34,
|
| 379 |
-
"learning_rate": 4.1654102950678443e-05,
|
| 380 |
-
"loss": 1.5704,
|
| 381 |
-
"step": 31000
|
| 382 |
-
},
|
| 383 |
-
{
|
| 384 |
-
"epoch": 3.39,
|
| 385 |
-
"learning_rate": 4.151949170794745e-05,
|
| 386 |
-
"loss": 1.5954,
|
| 387 |
-
"step": 31500
|
| 388 |
-
},
|
| 389 |
-
{
|
| 390 |
-
"epoch": 3.45,
|
| 391 |
-
"learning_rate": 4.138488046521646e-05,
|
| 392 |
-
"loss": 1.5908,
|
| 393 |
-
"step": 32000
|
| 394 |
-
},
|
| 395 |
-
{
|
| 396 |
-
"epoch": 3.5,
|
| 397 |
-
"learning_rate": 4.1250269222485464e-05,
|
| 398 |
-
"loss": 1.6013,
|
| 399 |
-
"step": 32500
|
| 400 |
-
},
|
| 401 |
-
{
|
| 402 |
-
"epoch": 3.55,
|
| 403 |
-
"learning_rate": 4.111565797975447e-05,
|
| 404 |
-
"loss": 1.6179,
|
| 405 |
-
"step": 33000
|
| 406 |
-
},
|
| 407 |
-
{
|
| 408 |
-
"epoch": 3.61,
|
| 409 |
-
"learning_rate": 4.098104673702348e-05,
|
| 410 |
-
"loss": 1.6316,
|
| 411 |
-
"step": 33500
|
| 412 |
-
},
|
| 413 |
-
{
|
| 414 |
-
"epoch": 3.66,
|
| 415 |
-
"learning_rate": 4.0846435494292485e-05,
|
| 416 |
-
"loss": 1.6168,
|
| 417 |
-
"step": 34000
|
| 418 |
-
},
|
| 419 |
-
{
|
| 420 |
-
"epoch": 3.72,
|
| 421 |
-
"learning_rate": 4.071182425156149e-05,
|
| 422 |
-
"loss": 1.6459,
|
| 423 |
-
"step": 34500
|
| 424 |
-
},
|
| 425 |
-
{
|
| 426 |
-
"epoch": 3.77,
|
| 427 |
-
"learning_rate": 4.05772130088305e-05,
|
| 428 |
-
"loss": 1.6441,
|
| 429 |
-
"step": 35000
|
| 430 |
-
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 3.82,
|
| 433 |
-
"learning_rate": 4.0442601766099506e-05,
|
| 434 |
-
"loss": 1.6658,
|
| 435 |
-
"step": 35500
|
| 436 |
-
},
|
| 437 |
-
{
|
| 438 |
-
"epoch": 3.88,
|
| 439 |
-
"learning_rate": 4.030799052336852e-05,
|
| 440 |
-
"loss": 1.6688,
|
| 441 |
-
"step": 36000
|
| 442 |
-
},
|
| 443 |
-
{
|
| 444 |
-
"epoch": 3.93,
|
| 445 |
-
"learning_rate": 4.017337928063752e-05,
|
| 446 |
-
"loss": 1.6433,
|
| 447 |
-
"step": 36500
|
| 448 |
-
},
|
| 449 |
-
{
|
| 450 |
-
"epoch": 3.98,
|
| 451 |
-
"learning_rate": 4.0038768037906526e-05,
|
| 452 |
-
"loss": 1.645,
|
| 453 |
-
"step": 37000
|
| 454 |
-
},
|
| 455 |
-
{
|
| 456 |
-
"epoch": 4.04,
|
| 457 |
-
"learning_rate": 3.990415679517553e-05,
|
| 458 |
-
"loss": 1.461,
|
| 459 |
-
"step": 37500
|
| 460 |
-
},
|
| 461 |
-
{
|
| 462 |
-
"epoch": 4.09,
|
| 463 |
-
"learning_rate": 3.976954555244454e-05,
|
| 464 |
-
"loss": 1.4231,
|
| 465 |
-
"step": 38000
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 4.15,
|
| 469 |
-
"learning_rate": 3.9634934309713554e-05,
|
| 470 |
-
"loss": 1.4397,
|
| 471 |
-
"step": 38500
|
| 472 |
-
},
|
| 473 |
-
{
|
| 474 |
-
"epoch": 4.2,
|
| 475 |
-
"learning_rate": 3.950032306698256e-05,
|
| 476 |
-
"loss": 1.4422,
|
| 477 |
-
"step": 39000
|
| 478 |
-
},
|
| 479 |
-
{
|
| 480 |
-
"epoch": 4.25,
|
| 481 |
-
"learning_rate": 3.936571182425156e-05,
|
| 482 |
-
"loss": 1.4731,
|
| 483 |
-
"step": 39500
|
| 484 |
-
},
|
| 485 |
-
{
|
| 486 |
-
"epoch": 4.31,
|
| 487 |
-
"learning_rate": 3.923110058152057e-05,
|
| 488 |
-
"loss": 1.4869,
|
| 489 |
-
"step": 40000
|
| 490 |
-
},
|
| 491 |
-
{
|
| 492 |
-
"epoch": 4.36,
|
| 493 |
-
"learning_rate": 3.9096489338789575e-05,
|
| 494 |
-
"loss": 1.4939,
|
| 495 |
-
"step": 40500
|
| 496 |
-
},
|
| 497 |
-
{
|
| 498 |
-
"epoch": 4.42,
|
| 499 |
-
"learning_rate": 3.896187809605859e-05,
|
| 500 |
-
"loss": 1.5142,
|
| 501 |
-
"step": 41000
|
| 502 |
-
},
|
| 503 |
-
{
|
| 504 |
-
"epoch": 4.47,
|
| 505 |
-
"learning_rate": 3.8827266853327595e-05,
|
| 506 |
-
"loss": 1.4907,
|
| 507 |
-
"step": 41500
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 4.52,
|
| 511 |
-
"learning_rate": 3.86926556105966e-05,
|
| 512 |
-
"loss": 1.5144,
|
| 513 |
-
"step": 42000
|
| 514 |
-
},
|
| 515 |
-
{
|
| 516 |
-
"epoch": 4.58,
|
| 517 |
-
"learning_rate": 3.85580443678656e-05,
|
| 518 |
-
"loss": 1.5211,
|
| 519 |
-
"step": 42500
|
| 520 |
-
},
|
| 521 |
-
{
|
| 522 |
-
"epoch": 4.63,
|
| 523 |
-
"learning_rate": 3.842343312513461e-05,
|
| 524 |
-
"loss": 1.5204,
|
| 525 |
-
"step": 43000
|
| 526 |
-
},
|
| 527 |
-
{
|
| 528 |
-
"epoch": 4.68,
|
| 529 |
-
"learning_rate": 3.828882188240362e-05,
|
| 530 |
-
"loss": 1.5474,
|
| 531 |
-
"step": 43500
|
| 532 |
-
},
|
| 533 |
-
{
|
| 534 |
-
"epoch": 4.74,
|
| 535 |
-
"learning_rate": 3.815421063967263e-05,
|
| 536 |
-
"loss": 1.5521,
|
| 537 |
-
"step": 44000
|
| 538 |
-
},
|
| 539 |
-
{
|
| 540 |
-
"epoch": 4.79,
|
| 541 |
-
"learning_rate": 3.8019599396941636e-05,
|
| 542 |
-
"loss": 1.5533,
|
| 543 |
-
"step": 44500
|
| 544 |
-
},
|
| 545 |
-
{
|
| 546 |
-
"epoch": 4.85,
|
| 547 |
-
"learning_rate": 3.788498815421064e-05,
|
| 548 |
-
"loss": 1.5394,
|
| 549 |
-
"step": 45000
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 4.9,
|
| 553 |
-
"learning_rate": 3.775037691147965e-05,
|
| 554 |
-
"loss": 1.5555,
|
| 555 |
-
"step": 45500
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"epoch": 4.95,
|
| 559 |
-
"learning_rate": 3.761576566874865e-05,
|
| 560 |
-
"loss": 1.5908,
|
| 561 |
-
"step": 46000
|
| 562 |
-
},
|
| 563 |
-
{
|
| 564 |
-
"epoch": 5.01,
|
| 565 |
-
"learning_rate": 3.7481154426017664e-05,
|
| 566 |
-
"loss": 1.5367,
|
| 567 |
-
"step": 46500
|
| 568 |
-
},
|
| 569 |
-
{
|
| 570 |
-
"epoch": 5.06,
|
| 571 |
-
"learning_rate": 3.734654318328667e-05,
|
| 572 |
-
"loss": 1.3431,
|
| 573 |
-
"step": 47000
|
| 574 |
-
},
|
| 575 |
-
{
|
| 576 |
-
"epoch": 5.12,
|
| 577 |
-
"learning_rate": 3.721193194055568e-05,
|
| 578 |
-
"loss": 1.3472,
|
| 579 |
-
"step": 47500
|
| 580 |
-
},
|
| 581 |
-
{
|
| 582 |
-
"epoch": 5.17,
|
| 583 |
-
"learning_rate": 3.7077320697824685e-05,
|
| 584 |
-
"loss": 1.3687,
|
| 585 |
-
"step": 48000
|
| 586 |
-
},
|
| 587 |
-
{
|
| 588 |
-
"epoch": 5.22,
|
| 589 |
-
"learning_rate": 3.694270945509369e-05,
|
| 590 |
-
"loss": 1.3831,
|
| 591 |
-
"step": 48500
|
| 592 |
-
},
|
| 593 |
-
{
|
| 594 |
-
"epoch": 5.28,
|
| 595 |
-
"learning_rate": 3.68080982123627e-05,
|
| 596 |
-
"loss": 1.3774,
|
| 597 |
-
"step": 49000
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 5.33,
|
| 601 |
-
"learning_rate": 3.6673486969631705e-05,
|
| 602 |
-
"loss": 1.3933,
|
| 603 |
-
"step": 49500
|
| 604 |
-
},
|
| 605 |
-
{
|
| 606 |
-
"epoch": 5.38,
|
| 607 |
-
"learning_rate": 3.653887572690071e-05,
|
| 608 |
-
"loss": 1.3961,
|
| 609 |
-
"step": 50000
|
| 610 |
-
},
|
| 611 |
-
{
|
| 612 |
-
"epoch": 5.44,
|
| 613 |
-
"learning_rate": 3.640426448416972e-05,
|
| 614 |
-
"loss": 1.4055,
|
| 615 |
-
"step": 50500
|
| 616 |
-
},
|
| 617 |
-
{
|
| 618 |
-
"epoch": 5.49,
|
| 619 |
-
"learning_rate": 3.6269653241438726e-05,
|
| 620 |
-
"loss": 1.4462,
|
| 621 |
-
"step": 51000
|
| 622 |
-
},
|
| 623 |
-
{
|
| 624 |
-
"epoch": 5.55,
|
| 625 |
-
"learning_rate": 3.613504199870773e-05,
|
| 626 |
-
"loss": 1.4235,
|
| 627 |
-
"step": 51500
|
| 628 |
-
},
|
| 629 |
-
{
|
| 630 |
-
"epoch": 5.6,
|
| 631 |
-
"learning_rate": 3.600043075597674e-05,
|
| 632 |
-
"loss": 1.4611,
|
| 633 |
-
"step": 52000
|
| 634 |
-
},
|
| 635 |
-
{
|
| 636 |
-
"epoch": 5.65,
|
| 637 |
-
"learning_rate": 3.586581951324575e-05,
|
| 638 |
-
"loss": 1.4476,
|
| 639 |
-
"step": 52500
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 5.71,
|
| 643 |
-
"learning_rate": 3.5731208270514754e-05,
|
| 644 |
-
"loss": 1.4708,
|
| 645 |
-
"step": 53000
|
| 646 |
-
},
|
| 647 |
-
{
|
| 648 |
-
"epoch": 5.76,
|
| 649 |
-
"learning_rate": 3.559659702778376e-05,
|
| 650 |
-
"loss": 1.4818,
|
| 651 |
-
"step": 53500
|
| 652 |
-
},
|
| 653 |
-
{
|
| 654 |
-
"epoch": 5.82,
|
| 655 |
-
"learning_rate": 3.546198578505277e-05,
|
| 656 |
-
"loss": 1.4486,
|
| 657 |
-
"step": 54000
|
| 658 |
-
},
|
| 659 |
-
{
|
| 660 |
-
"epoch": 5.87,
|
| 661 |
-
"learning_rate": 3.532737454232178e-05,
|
| 662 |
-
"loss": 1.466,
|
| 663 |
-
"step": 54500
|
| 664 |
-
},
|
| 665 |
-
{
|
| 666 |
-
"epoch": 5.92,
|
| 667 |
-
"learning_rate": 3.519276329959078e-05,
|
| 668 |
-
"loss": 1.4813,
|
| 669 |
-
"step": 55000
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 5.98,
|
| 673 |
-
"learning_rate": 3.505815205685979e-05,
|
| 674 |
-
"loss": 1.4932,
|
| 675 |
-
"step": 55500
|
| 676 |
-
},
|
| 677 |
-
{
|
| 678 |
-
"epoch": 6.03,
|
| 679 |
-
"learning_rate": 3.4923540814128795e-05,
|
| 680 |
-
"loss": 1.3391,
|
| 681 |
-
"step": 56000
|
| 682 |
-
},
|
| 683 |
-
{
|
| 684 |
-
"epoch": 6.08,
|
| 685 |
-
"learning_rate": 3.47889295713978e-05,
|
| 686 |
-
"loss": 1.2801,
|
| 687 |
-
"step": 56500
|
| 688 |
-
},
|
| 689 |
-
{
|
| 690 |
-
"epoch": 6.14,
|
| 691 |
-
"learning_rate": 3.4654318328666816e-05,
|
| 692 |
-
"loss": 1.2803,
|
| 693 |
-
"step": 57000
|
| 694 |
-
},
|
| 695 |
-
{
|
| 696 |
-
"epoch": 6.19,
|
| 697 |
-
"learning_rate": 3.451970708593582e-05,
|
| 698 |
-
"loss": 1.2791,
|
| 699 |
-
"step": 57500
|
| 700 |
-
},
|
| 701 |
-
{
|
| 702 |
-
"epoch": 6.25,
|
| 703 |
-
"learning_rate": 3.438509584320483e-05,
|
| 704 |
-
"loss": 1.297,
|
| 705 |
-
"step": 58000
|
| 706 |
-
},
|
| 707 |
-
{
|
| 708 |
-
"epoch": 6.3,
|
| 709 |
-
"learning_rate": 3.425048460047383e-05,
|
| 710 |
-
"loss": 1.304,
|
| 711 |
-
"step": 58500
|
| 712 |
-
},
|
| 713 |
-
{
|
| 714 |
-
"epoch": 6.35,
|
| 715 |
-
"learning_rate": 3.4115873357742836e-05,
|
| 716 |
-
"loss": 1.3166,
|
| 717 |
-
"step": 59000
|
| 718 |
-
},
|
| 719 |
-
{
|
| 720 |
-
"epoch": 6.41,
|
| 721 |
-
"learning_rate": 3.398126211501185e-05,
|
| 722 |
-
"loss": 1.3318,
|
| 723 |
-
"step": 59500
|
| 724 |
-
},
|
| 725 |
-
{
|
| 726 |
-
"epoch": 6.46,
|
| 727 |
-
"learning_rate": 3.384665087228086e-05,
|
| 728 |
-
"loss": 1.3501,
|
| 729 |
-
"step": 60000
|
| 730 |
-
},
|
| 731 |
-
{
|
| 732 |
-
"epoch": 6.52,
|
| 733 |
-
"learning_rate": 3.3712039629549864e-05,
|
| 734 |
-
"loss": 1.3576,
|
| 735 |
-
"step": 60500
|
| 736 |
-
},
|
| 737 |
-
{
|
| 738 |
-
"epoch": 6.57,
|
| 739 |
-
"learning_rate": 3.357742838681887e-05,
|
| 740 |
-
"loss": 1.3679,
|
| 741 |
-
"step": 61000
|
| 742 |
-
},
|
| 743 |
-
{
|
| 744 |
-
"epoch": 6.62,
|
| 745 |
-
"learning_rate": 3.344281714408787e-05,
|
| 746 |
-
"loss": 1.3905,
|
| 747 |
-
"step": 61500
|
| 748 |
-
},
|
| 749 |
-
{
|
| 750 |
-
"epoch": 6.68,
|
| 751 |
-
"learning_rate": 3.3308205901356885e-05,
|
| 752 |
-
"loss": 1.361,
|
| 753 |
-
"step": 62000
|
| 754 |
-
},
|
| 755 |
-
{
|
| 756 |
-
"epoch": 6.73,
|
| 757 |
-
"learning_rate": 3.317359465862589e-05,
|
| 758 |
-
"loss": 1.3657,
|
| 759 |
-
"step": 62500
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 6.78,
|
| 763 |
-
"learning_rate": 3.30389834158949e-05,
|
| 764 |
-
"loss": 1.3768,
|
| 765 |
-
"step": 63000
|
| 766 |
-
},
|
| 767 |
-
{
|
| 768 |
-
"epoch": 6.84,
|
| 769 |
-
"learning_rate": 3.2904372173163905e-05,
|
| 770 |
-
"loss": 1.4069,
|
| 771 |
-
"step": 63500
|
| 772 |
-
},
|
| 773 |
-
{
|
| 774 |
-
"epoch": 6.89,
|
| 775 |
-
"learning_rate": 3.276976093043291e-05,
|
| 776 |
-
"loss": 1.3913,
|
| 777 |
-
"step": 64000
|
| 778 |
-
},
|
| 779 |
-
{
|
| 780 |
-
"epoch": 6.95,
|
| 781 |
-
"learning_rate": 3.263514968770192e-05,
|
| 782 |
-
"loss": 1.3974,
|
| 783 |
-
"step": 64500
|
| 784 |
-
},
|
| 785 |
-
{
|
| 786 |
-
"epoch": 7.0,
|
| 787 |
-
"learning_rate": 3.2500538444970926e-05,
|
| 788 |
-
"loss": 1.4024,
|
| 789 |
-
"step": 65000
|
| 790 |
-
},
|
| 791 |
-
{
|
| 792 |
-
"epoch": 7.05,
|
| 793 |
-
"learning_rate": 3.236592720223993e-05,
|
| 794 |
-
"loss": 1.1818,
|
| 795 |
-
"step": 65500
|
| 796 |
-
},
|
| 797 |
-
{
|
| 798 |
-
"epoch": 7.11,
|
| 799 |
-
"learning_rate": 3.223131595950894e-05,
|
| 800 |
-
"loss": 1.2059,
|
| 801 |
-
"step": 66000
|
| 802 |
-
},
|
| 803 |
-
{
|
| 804 |
-
"epoch": 7.16,
|
| 805 |
-
"learning_rate": 3.209670471677795e-05,
|
| 806 |
-
"loss": 1.2021,
|
| 807 |
-
"step": 66500
|
| 808 |
-
},
|
| 809 |
-
{
|
| 810 |
-
"epoch": 7.22,
|
| 811 |
-
"learning_rate": 3.1962093474046954e-05,
|
| 812 |
-
"loss": 1.2224,
|
| 813 |
-
"step": 67000
|
| 814 |
-
},
|
| 815 |
-
{
|
| 816 |
-
"epoch": 7.27,
|
| 817 |
-
"learning_rate": 3.182748223131596e-05,
|
| 818 |
-
"loss": 1.2637,
|
| 819 |
-
"step": 67500
|
| 820 |
-
},
|
| 821 |
-
{
|
| 822 |
-
"epoch": 7.32,
|
| 823 |
-
"learning_rate": 3.169287098858497e-05,
|
| 824 |
-
"loss": 1.246,
|
| 825 |
-
"step": 68000
|
| 826 |
-
},
|
| 827 |
-
{
|
| 828 |
-
"epoch": 7.38,
|
| 829 |
-
"learning_rate": 3.1558259745853974e-05,
|
| 830 |
-
"loss": 1.2506,
|
| 831 |
-
"step": 68500
|
| 832 |
-
},
|
| 833 |
-
{
|
| 834 |
-
"epoch": 7.43,
|
| 835 |
-
"learning_rate": 3.142364850312298e-05,
|
| 836 |
-
"loss": 1.2749,
|
| 837 |
-
"step": 69000
|
| 838 |
-
},
|
| 839 |
-
{
|
| 840 |
-
"epoch": 7.48,
|
| 841 |
-
"learning_rate": 3.128903726039199e-05,
|
| 842 |
-
"loss": 1.2853,
|
| 843 |
-
"step": 69500
|
| 844 |
-
},
|
| 845 |
-
{
|
| 846 |
-
"epoch": 7.54,
|
| 847 |
-
"learning_rate": 3.1154426017660995e-05,
|
| 848 |
-
"loss": 1.2919,
|
| 849 |
-
"step": 70000
|
| 850 |
-
},
|
| 851 |
-
{
|
| 852 |
-
"epoch": 7.59,
|
| 853 |
-
"learning_rate": 3.101981477493001e-05,
|
| 854 |
-
"loss": 1.307,
|
| 855 |
-
"step": 70500
|
| 856 |
-
},
|
| 857 |
-
{
|
| 858 |
-
"epoch": 7.65,
|
| 859 |
-
"learning_rate": 3.088520353219901e-05,
|
| 860 |
-
"loss": 1.3098,
|
| 861 |
-
"step": 71000
|
| 862 |
-
},
|
| 863 |
-
{
|
| 864 |
-
"epoch": 7.7,
|
| 865 |
-
"learning_rate": 3.0750592289468016e-05,
|
| 866 |
-
"loss": 1.2979,
|
| 867 |
-
"step": 71500
|
| 868 |
-
},
|
| 869 |
-
{
|
| 870 |
-
"epoch": 7.75,
|
| 871 |
-
"learning_rate": 3.061598104673702e-05,
|
| 872 |
-
"loss": 1.3009,
|
| 873 |
-
"step": 72000
|
| 874 |
-
},
|
| 875 |
-
{
|
| 876 |
-
"epoch": 7.81,
|
| 877 |
-
"learning_rate": 3.0481369804006033e-05,
|
| 878 |
-
"loss": 1.2957,
|
| 879 |
-
"step": 72500
|
| 880 |
-
},
|
| 881 |
-
{
|
| 882 |
-
"epoch": 7.86,
|
| 883 |
-
"learning_rate": 3.034675856127504e-05,
|
| 884 |
-
"loss": 1.3204,
|
| 885 |
-
"step": 73000
|
| 886 |
-
},
|
| 887 |
-
{
|
| 888 |
-
"epoch": 7.92,
|
| 889 |
-
"learning_rate": 3.021214731854405e-05,
|
| 890 |
-
"loss": 1.3106,
|
| 891 |
-
"step": 73500
|
| 892 |
-
},
|
| 893 |
-
{
|
| 894 |
-
"epoch": 7.97,
|
| 895 |
-
"learning_rate": 3.007753607581305e-05,
|
| 896 |
-
"loss": 1.3095,
|
| 897 |
-
"step": 74000
|
| 898 |
-
},
|
| 899 |
-
{
|
| 900 |
-
"epoch": 8.02,
|
| 901 |
-
"learning_rate": 2.994292483308206e-05,
|
| 902 |
-
"loss": 1.214,
|
| 903 |
-
"step": 74500
|
| 904 |
-
},
|
| 905 |
-
{
|
| 906 |
-
"epoch": 8.08,
|
| 907 |
-
"learning_rate": 2.9808313590351067e-05,
|
| 908 |
-
"loss": 1.1334,
|
| 909 |
-
"step": 75000
|
| 910 |
-
},
|
| 911 |
-
{
|
| 912 |
-
"epoch": 8.13,
|
| 913 |
-
"learning_rate": 2.9673702347620074e-05,
|
| 914 |
-
"loss": 1.1408,
|
| 915 |
-
"step": 75500
|
| 916 |
-
},
|
| 917 |
-
{
|
| 918 |
-
"epoch": 8.18,
|
| 919 |
-
"learning_rate": 2.9539091104889084e-05,
|
| 920 |
-
"loss": 1.151,
|
| 921 |
-
"step": 76000
|
| 922 |
-
},
|
| 923 |
-
{
|
| 924 |
-
"epoch": 8.24,
|
| 925 |
-
"learning_rate": 2.940447986215809e-05,
|
| 926 |
-
"loss": 1.1552,
|
| 927 |
-
"step": 76500
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 8.29,
|
| 931 |
-
"learning_rate": 2.9269868619427095e-05,
|
| 932 |
-
"loss": 1.1694,
|
| 933 |
-
"step": 77000
|
| 934 |
-
},
|
| 935 |
-
{
|
| 936 |
-
"epoch": 8.35,
|
| 937 |
-
"learning_rate": 2.9135257376696102e-05,
|
| 938 |
-
"loss": 1.193,
|
| 939 |
-
"step": 77500
|
| 940 |
-
},
|
| 941 |
-
{
|
| 942 |
-
"epoch": 8.4,
|
| 943 |
-
"learning_rate": 2.900064613396511e-05,
|
| 944 |
-
"loss": 1.1861,
|
| 945 |
-
"step": 78000
|
| 946 |
-
},
|
| 947 |
-
{
|
| 948 |
-
"epoch": 8.45,
|
| 949 |
-
"learning_rate": 2.8866034891234116e-05,
|
| 950 |
-
"loss": 1.1871,
|
| 951 |
-
"step": 78500
|
| 952 |
-
},
|
| 953 |
-
{
|
| 954 |
-
"epoch": 8.51,
|
| 955 |
-
"learning_rate": 2.8731423648503126e-05,
|
| 956 |
-
"loss": 1.2038,
|
| 957 |
-
"step": 79000
|
| 958 |
-
},
|
| 959 |
-
{
|
| 960 |
-
"epoch": 8.56,
|
| 961 |
-
"learning_rate": 2.8596812405772133e-05,
|
| 962 |
-
"loss": 1.2259,
|
| 963 |
-
"step": 79500
|
| 964 |
-
},
|
| 965 |
-
{
|
| 966 |
-
"epoch": 8.62,
|
| 967 |
-
"learning_rate": 2.846220116304114e-05,
|
| 968 |
-
"loss": 1.2174,
|
| 969 |
-
"step": 80000
|
| 970 |
-
},
|
| 971 |
-
{
|
| 972 |
-
"epoch": 8.67,
|
| 973 |
-
"learning_rate": 2.8327589920310143e-05,
|
| 974 |
-
"loss": 1.2217,
|
| 975 |
-
"step": 80500
|
| 976 |
-
},
|
| 977 |
-
{
|
| 978 |
-
"epoch": 8.72,
|
| 979 |
-
"learning_rate": 2.819297867757915e-05,
|
| 980 |
-
"loss": 1.2326,
|
| 981 |
-
"step": 81000
|
| 982 |
-
},
|
| 983 |
-
{
|
| 984 |
-
"epoch": 8.78,
|
| 985 |
-
"learning_rate": 2.805836743484816e-05,
|
| 986 |
-
"loss": 1.2494,
|
| 987 |
-
"step": 81500
|
| 988 |
-
},
|
| 989 |
-
{
|
| 990 |
-
"epoch": 8.83,
|
| 991 |
-
"learning_rate": 2.7923756192117167e-05,
|
| 992 |
-
"loss": 1.2486,
|
| 993 |
-
"step": 82000
|
| 994 |
-
},
|
| 995 |
-
{
|
| 996 |
-
"epoch": 8.88,
|
| 997 |
-
"learning_rate": 2.7789144949386174e-05,
|
| 998 |
-
"loss": 1.2447,
|
| 999 |
-
"step": 82500
|
| 1000 |
-
},
|
| 1001 |
-
{
|
| 1002 |
-
"epoch": 8.94,
|
| 1003 |
-
"learning_rate": 2.7654533706655184e-05,
|
| 1004 |
-
"loss": 1.2425,
|
| 1005 |
-
"step": 83000
|
| 1006 |
-
},
|
| 1007 |
-
{
|
| 1008 |
-
"epoch": 8.99,
|
| 1009 |
-
"learning_rate": 2.7519922463924185e-05,
|
| 1010 |
-
"loss": 1.258,
|
| 1011 |
-
"step": 83500
|
| 1012 |
-
},
|
| 1013 |
-
{
|
| 1014 |
-
"epoch": 9.05,
|
| 1015 |
-
"learning_rate": 2.7385311221193195e-05,
|
| 1016 |
-
"loss": 1.0903,
|
| 1017 |
-
"step": 84000
|
| 1018 |
-
},
|
| 1019 |
-
{
|
| 1020 |
-
"epoch": 9.1,
|
| 1021 |
-
"learning_rate": 2.7250699978462202e-05,
|
| 1022 |
-
"loss": 1.0811,
|
| 1023 |
-
"step": 84500
|
| 1024 |
-
},
|
| 1025 |
-
{
|
| 1026 |
-
"epoch": 9.15,
|
| 1027 |
-
"learning_rate": 2.711608873573121e-05,
|
| 1028 |
-
"loss": 1.0894,
|
| 1029 |
-
"step": 85000
|
| 1030 |
-
},
|
| 1031 |
-
{
|
| 1032 |
-
"epoch": 9.21,
|
| 1033 |
-
"learning_rate": 2.698147749300022e-05,
|
| 1034 |
-
"loss": 1.1093,
|
| 1035 |
-
"step": 85500
|
| 1036 |
-
},
|
| 1037 |
-
{
|
| 1038 |
-
"epoch": 9.26,
|
| 1039 |
-
"learning_rate": 2.6846866250269226e-05,
|
| 1040 |
-
"loss": 1.0984,
|
| 1041 |
-
"step": 86000
|
| 1042 |
-
},
|
| 1043 |
-
{
|
| 1044 |
-
"epoch": 9.32,
|
| 1045 |
-
"learning_rate": 2.671225500753823e-05,
|
| 1046 |
-
"loss": 1.0906,
|
| 1047 |
-
"step": 86500
|
| 1048 |
-
},
|
| 1049 |
-
{
|
| 1050 |
-
"epoch": 9.37,
|
| 1051 |
-
"learning_rate": 2.6577643764807236e-05,
|
| 1052 |
-
"loss": 1.109,
|
| 1053 |
-
"step": 87000
|
| 1054 |
-
},
|
| 1055 |
-
{
|
| 1056 |
-
"epoch": 9.42,
|
| 1057 |
-
"learning_rate": 2.6443032522076243e-05,
|
| 1058 |
-
"loss": 1.1383,
|
| 1059 |
-
"step": 87500
|
| 1060 |
-
},
|
| 1061 |
-
{
|
| 1062 |
-
"epoch": 9.48,
|
| 1063 |
-
"learning_rate": 2.6308421279345253e-05,
|
| 1064 |
-
"loss": 1.146,
|
| 1065 |
-
"step": 88000
|
| 1066 |
-
},
|
| 1067 |
-
{
|
| 1068 |
-
"epoch": 9.53,
|
| 1069 |
-
"learning_rate": 2.617381003661426e-05,
|
| 1070 |
-
"loss": 1.1484,
|
| 1071 |
-
"step": 88500
|
| 1072 |
-
},
|
| 1073 |
-
{
|
| 1074 |
-
"epoch": 9.58,
|
| 1075 |
-
"learning_rate": 2.6039198793883267e-05,
|
| 1076 |
-
"loss": 1.1512,
|
| 1077 |
-
"step": 89000
|
| 1078 |
-
},
|
| 1079 |
-
{
|
| 1080 |
-
"epoch": 9.64,
|
| 1081 |
-
"learning_rate": 2.5904587551152277e-05,
|
| 1082 |
-
"loss": 1.154,
|
| 1083 |
-
"step": 89500
|
| 1084 |
-
},
|
| 1085 |
-
{
|
| 1086 |
-
"epoch": 9.69,
|
| 1087 |
-
"learning_rate": 2.5769976308421278e-05,
|
| 1088 |
-
"loss": 1.1533,
|
| 1089 |
-
"step": 90000
|
| 1090 |
-
},
|
| 1091 |
-
{
|
| 1092 |
-
"epoch": 9.75,
|
| 1093 |
-
"learning_rate": 2.5635365065690288e-05,
|
| 1094 |
-
"loss": 1.183,
|
| 1095 |
-
"step": 90500
|
| 1096 |
-
},
|
| 1097 |
-
{
|
| 1098 |
-
"epoch": 9.8,
|
| 1099 |
-
"learning_rate": 2.5500753822959295e-05,
|
| 1100 |
-
"loss": 1.1628,
|
| 1101 |
-
"step": 91000
|
| 1102 |
-
},
|
| 1103 |
-
{
|
| 1104 |
-
"epoch": 9.85,
|
| 1105 |
-
"learning_rate": 2.53661425802283e-05,
|
| 1106 |
-
"loss": 1.1586,
|
| 1107 |
-
"step": 91500
|
| 1108 |
-
},
|
| 1109 |
-
{
|
| 1110 |
-
"epoch": 9.91,
|
| 1111 |
-
"learning_rate": 2.5231531337497312e-05,
|
| 1112 |
-
"loss": 1.1968,
|
| 1113 |
-
"step": 92000
|
| 1114 |
-
},
|
| 1115 |
-
{
|
| 1116 |
-
"epoch": 9.96,
|
| 1117 |
-
"learning_rate": 2.509692009476632e-05,
|
| 1118 |
-
"loss": 1.1853,
|
| 1119 |
-
"step": 92500
|
| 1120 |
-
},
|
| 1121 |
-
{
|
| 1122 |
-
"epoch": 10.02,
|
| 1123 |
-
"learning_rate": 2.4962308852035322e-05,
|
| 1124 |
-
"loss": 1.1338,
|
| 1125 |
-
"step": 93000
|
| 1126 |
-
},
|
| 1127 |
-
{
|
| 1128 |
-
"epoch": 10.07,
|
| 1129 |
-
"learning_rate": 2.482769760930433e-05,
|
| 1130 |
-
"loss": 1.0121,
|
| 1131 |
-
"step": 93500
|
| 1132 |
-
},
|
| 1133 |
-
{
|
| 1134 |
-
"epoch": 10.12,
|
| 1135 |
-
"learning_rate": 2.4693086366573336e-05,
|
| 1136 |
-
"loss": 1.0293,
|
| 1137 |
-
"step": 94000
|
| 1138 |
-
},
|
| 1139 |
-
{
|
| 1140 |
-
"epoch": 10.18,
|
| 1141 |
-
"learning_rate": 2.4558475123842346e-05,
|
| 1142 |
-
"loss": 1.0306,
|
| 1143 |
-
"step": 94500
|
| 1144 |
-
},
|
| 1145 |
-
{
|
| 1146 |
-
"epoch": 10.23,
|
| 1147 |
-
"learning_rate": 2.4423863881111353e-05,
|
| 1148 |
-
"loss": 1.0396,
|
| 1149 |
-
"step": 95000
|
| 1150 |
-
},
|
| 1151 |
-
{
|
| 1152 |
-
"epoch": 10.28,
|
| 1153 |
-
"learning_rate": 2.4289252638380357e-05,
|
| 1154 |
-
"loss": 1.0501,
|
| 1155 |
-
"step": 95500
|
| 1156 |
-
},
|
| 1157 |
-
{
|
| 1158 |
-
"epoch": 10.34,
|
| 1159 |
-
"learning_rate": 2.4154641395649367e-05,
|
| 1160 |
-
"loss": 1.0499,
|
| 1161 |
-
"step": 96000
|
| 1162 |
-
},
|
| 1163 |
-
{
|
| 1164 |
-
"epoch": 10.39,
|
| 1165 |
-
"learning_rate": 2.4020030152918374e-05,
|
| 1166 |
-
"loss": 1.0542,
|
| 1167 |
-
"step": 96500
|
| 1168 |
-
},
|
| 1169 |
-
{
|
| 1170 |
-
"epoch": 10.45,
|
| 1171 |
-
"learning_rate": 2.388541891018738e-05,
|
| 1172 |
-
"loss": 1.0756,
|
| 1173 |
-
"step": 97000
|
| 1174 |
-
},
|
| 1175 |
-
{
|
| 1176 |
-
"epoch": 10.5,
|
| 1177 |
-
"learning_rate": 2.3750807667456388e-05,
|
| 1178 |
-
"loss": 1.0793,
|
| 1179 |
-
"step": 97500
|
| 1180 |
-
},
|
| 1181 |
-
{
|
| 1182 |
-
"epoch": 10.55,
|
| 1183 |
-
"learning_rate": 2.3616196424725395e-05,
|
| 1184 |
-
"loss": 1.0988,
|
| 1185 |
-
"step": 98000
|
| 1186 |
-
},
|
| 1187 |
-
{
|
| 1188 |
-
"epoch": 10.61,
|
| 1189 |
-
"learning_rate": 2.34815851819944e-05,
|
| 1190 |
-
"loss": 1.0935,
|
| 1191 |
-
"step": 98500
|
| 1192 |
-
},
|
| 1193 |
-
{
|
| 1194 |
-
"epoch": 10.66,
|
| 1195 |
-
"learning_rate": 2.334697393926341e-05,
|
| 1196 |
-
"loss": 1.091,
|
| 1197 |
-
"step": 99000
|
| 1198 |
-
},
|
| 1199 |
-
{
|
| 1200 |
-
"epoch": 10.72,
|
| 1201 |
-
"learning_rate": 2.3212362696532415e-05,
|
| 1202 |
-
"loss": 1.1021,
|
| 1203 |
-
"step": 99500
|
| 1204 |
-
},
|
| 1205 |
-
{
|
| 1206 |
-
"epoch": 10.77,
|
| 1207 |
-
"learning_rate": 2.3077751453801422e-05,
|
| 1208 |
-
"loss": 1.1107,
|
| 1209 |
-
"step": 100000
|
| 1210 |
-
},
|
| 1211 |
-
{
|
| 1212 |
-
"epoch": 10.82,
|
| 1213 |
-
"learning_rate": 2.294314021107043e-05,
|
| 1214 |
-
"loss": 1.1004,
|
| 1215 |
-
"step": 100500
|
| 1216 |
-
},
|
| 1217 |
-
{
|
| 1218 |
-
"epoch": 10.88,
|
| 1219 |
-
"learning_rate": 2.2808528968339436e-05,
|
| 1220 |
-
"loss": 1.1234,
|
| 1221 |
-
"step": 101000
|
| 1222 |
-
},
|
| 1223 |
-
{
|
| 1224 |
-
"epoch": 10.93,
|
| 1225 |
-
"learning_rate": 2.2673917725608443e-05,
|
| 1226 |
-
"loss": 1.12,
|
| 1227 |
-
"step": 101500
|
| 1228 |
-
},
|
| 1229 |
-
{
|
| 1230 |
-
"epoch": 10.98,
|
| 1231 |
-
"learning_rate": 2.253930648287745e-05,
|
| 1232 |
-
"loss": 1.1295,
|
| 1233 |
-
"step": 102000
|
| 1234 |
-
},
|
| 1235 |
-
{
|
| 1236 |
-
"epoch": 11.04,
|
| 1237 |
-
"learning_rate": 2.240469524014646e-05,
|
| 1238 |
-
"loss": 1.0042,
|
| 1239 |
-
"step": 102500
|
| 1240 |
-
},
|
| 1241 |
-
{
|
| 1242 |
-
"epoch": 11.09,
|
| 1243 |
-
"learning_rate": 2.2270083997415464e-05,
|
| 1244 |
-
"loss": 0.9767,
|
| 1245 |
-
"step": 103000
|
| 1246 |
-
},
|
| 1247 |
-
{
|
| 1248 |
-
"epoch": 11.15,
|
| 1249 |
-
"learning_rate": 2.213547275468447e-05,
|
| 1250 |
-
"loss": 0.9776,
|
| 1251 |
-
"step": 103500
|
| 1252 |
-
},
|
| 1253 |
-
{
|
| 1254 |
-
"epoch": 11.2,
|
| 1255 |
-
"learning_rate": 2.200086151195348e-05,
|
| 1256 |
-
"loss": 0.9777,
|
| 1257 |
-
"step": 104000
|
| 1258 |
-
},
|
| 1259 |
-
{
|
| 1260 |
-
"epoch": 11.25,
|
| 1261 |
-
"learning_rate": 2.1866250269222484e-05,
|
| 1262 |
-
"loss": 0.9975,
|
| 1263 |
-
"step": 104500
|
| 1264 |
-
},
|
| 1265 |
-
{
|
| 1266 |
-
"epoch": 11.31,
|
| 1267 |
-
"learning_rate": 2.1731639026491495e-05,
|
| 1268 |
-
"loss": 1.0122,
|
| 1269 |
-
"step": 105000
|
| 1270 |
-
},
|
| 1271 |
-
{
|
| 1272 |
-
"epoch": 11.36,
|
| 1273 |
-
"learning_rate": 2.15970277837605e-05,
|
| 1274 |
-
"loss": 0.991,
|
| 1275 |
-
"step": 105500
|
| 1276 |
-
},
|
| 1277 |
-
{
|
| 1278 |
-
"epoch": 11.42,
|
| 1279 |
-
"learning_rate": 2.146241654102951e-05,
|
| 1280 |
-
"loss": 1.0095,
|
| 1281 |
-
"step": 106000
|
| 1282 |
-
},
|
| 1283 |
-
{
|
| 1284 |
-
"epoch": 11.47,
|
| 1285 |
-
"learning_rate": 2.1327805298298515e-05,
|
| 1286 |
-
"loss": 1.0179,
|
| 1287 |
-
"step": 106500
|
| 1288 |
-
},
|
| 1289 |
-
{
|
| 1290 |
-
"epoch": 11.52,
|
| 1291 |
-
"learning_rate": 2.1193194055567522e-05,
|
| 1292 |
-
"loss": 1.019,
|
| 1293 |
-
"step": 107000
|
| 1294 |
-
},
|
| 1295 |
-
{
|
| 1296 |
-
"epoch": 11.58,
|
| 1297 |
-
"learning_rate": 2.105858281283653e-05,
|
| 1298 |
-
"loss": 1.0495,
|
| 1299 |
-
"step": 107500
|
| 1300 |
-
},
|
| 1301 |
-
{
|
| 1302 |
-
"epoch": 11.63,
|
| 1303 |
-
"learning_rate": 2.0923971570105536e-05,
|
| 1304 |
-
"loss": 1.0266,
|
| 1305 |
-
"step": 108000
|
| 1306 |
-
},
|
| 1307 |
-
{
|
| 1308 |
-
"epoch": 11.68,
|
| 1309 |
-
"learning_rate": 2.0789360327374543e-05,
|
| 1310 |
-
"loss": 1.0422,
|
| 1311 |
-
"step": 108500
|
| 1312 |
-
},
|
| 1313 |
-
{
|
| 1314 |
-
"epoch": 11.74,
|
| 1315 |
-
"learning_rate": 2.0654749084643553e-05,
|
| 1316 |
-
"loss": 1.0383,
|
| 1317 |
-
"step": 109000
|
| 1318 |
-
},
|
| 1319 |
-
{
|
| 1320 |
-
"epoch": 11.79,
|
| 1321 |
-
"learning_rate": 2.0520137841912557e-05,
|
| 1322 |
-
"loss": 1.0624,
|
| 1323 |
-
"step": 109500
|
| 1324 |
-
},
|
| 1325 |
-
{
|
| 1326 |
-
"epoch": 11.85,
|
| 1327 |
-
"learning_rate": 2.0385526599181564e-05,
|
| 1328 |
-
"loss": 1.064,
|
| 1329 |
-
"step": 110000
|
| 1330 |
}
|
| 1331 |
],
|
| 1332 |
"logging_steps": 500,
|
| 1333 |
-
"max_steps":
|
| 1334 |
"num_input_tokens_seen": 0,
|
| 1335 |
-
"num_train_epochs":
|
| 1336 |
"save_steps": 10000,
|
| 1337 |
-
"total_flos":
|
| 1338 |
-
"train_batch_size":
|
| 1339 |
"trial_name": null,
|
| 1340 |
"trial_params": null
|
| 1341 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.1491510853212983,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 10000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.11,
|
| 13 |
+
"learning_rate": 4.820904076223225e-05,
|
| 14 |
+
"loss": 1.1625,
|
| 15 |
"step": 500
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.21,
|
| 19 |
+
"learning_rate": 4.641808152446451e-05,
|
| 20 |
+
"loss": 1.1918,
|
| 21 |
"step": 1000
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.32,
|
| 25 |
+
"learning_rate": 4.462712228669675e-05,
|
| 26 |
+
"loss": 1.2037,
|
| 27 |
"step": 1500
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.43,
|
| 31 |
+
"learning_rate": 4.283616304892901e-05,
|
| 32 |
+
"loss": 1.2352,
|
| 33 |
"step": 2000
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 0.54,
|
| 37 |
+
"learning_rate": 4.104520381116126e-05,
|
| 38 |
+
"loss": 1.2375,
|
| 39 |
"step": 2500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.64,
|
| 43 |
+
"learning_rate": 3.9254244573393514e-05,
|
| 44 |
+
"loss": 1.2305,
|
| 45 |
"step": 3000
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.75,
|
| 49 |
+
"learning_rate": 3.746328533562576e-05,
|
| 50 |
+
"loss": 1.2822,
|
| 51 |
"step": 3500
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.86,
|
| 55 |
+
"learning_rate": 3.567232609785802e-05,
|
| 56 |
+
"loss": 1.2733,
|
| 57 |
"step": 4000
|
| 58 |
},
|
| 59 |
{
|
| 60 |
+
"epoch": 0.97,
|
| 61 |
+
"learning_rate": 3.388136686009026e-05,
|
| 62 |
+
"loss": 1.2598,
|
| 63 |
"step": 4500
|
| 64 |
},
|
| 65 |
{
|
| 66 |
+
"epoch": 1.07,
|
| 67 |
+
"learning_rate": 3.209040762232252e-05,
|
| 68 |
+
"loss": 1.1006,
|
| 69 |
"step": 5000
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 1.18,
|
| 73 |
+
"learning_rate": 3.029944838455477e-05,
|
| 74 |
+
"loss": 1.0509,
|
| 75 |
"step": 5500
|
| 76 |
},
|
| 77 |
{
|
| 78 |
+
"epoch": 1.29,
|
| 79 |
+
"learning_rate": 2.850848914678702e-05,
|
| 80 |
+
"loss": 1.0549,
|
| 81 |
"step": 6000
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 1.4,
|
| 85 |
+
"learning_rate": 2.6717529909019275e-05,
|
| 86 |
+
"loss": 1.0917,
|
| 87 |
"step": 6500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 1.5,
|
| 91 |
+
"learning_rate": 2.4926570671251524e-05,
|
| 92 |
+
"loss": 1.1086,
|
| 93 |
"step": 7000
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 1.61,
|
| 97 |
+
"learning_rate": 2.3135611433483774e-05,
|
| 98 |
+
"loss": 1.0815,
|
| 99 |
"step": 7500
|
| 100 |
},
|
| 101 |
{
|
| 102 |
+
"epoch": 1.72,
|
| 103 |
+
"learning_rate": 2.1344652195716027e-05,
|
| 104 |
+
"loss": 1.0995,
|
| 105 |
"step": 8000
|
| 106 |
},
|
| 107 |
{
|
| 108 |
+
"epoch": 1.83,
|
| 109 |
+
"learning_rate": 1.9553692957948277e-05,
|
| 110 |
+
"loss": 1.1132,
|
| 111 |
"step": 8500
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 1.93,
|
| 115 |
+
"learning_rate": 1.776273372018053e-05,
|
| 116 |
+
"loss": 1.075,
|
| 117 |
"step": 9000
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 2.04,
|
| 121 |
+
"learning_rate": 1.5971774482412783e-05,
|
| 122 |
+
"loss": 1.054,
|
| 123 |
"step": 9500
|
| 124 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 125 |
{
|
| 126 |
"epoch": 2.15,
|
| 127 |
+
"learning_rate": 1.4180815244645032e-05,
|
| 128 |
+
"loss": 0.9583,
|
| 129 |
+
"step": 10000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 130 |
}
|
| 131 |
],
|
| 132 |
"logging_steps": 500,
|
| 133 |
+
"max_steps": 13959,
|
| 134 |
"num_input_tokens_seen": 0,
|
| 135 |
+
"num_train_epochs": 3,
|
| 136 |
"save_steps": 10000,
|
| 137 |
+
"total_flos": 2.0902317391872e+16,
|
| 138 |
+
"train_batch_size": 8,
|
| 139 |
"trial_name": null,
|
| 140 |
"trial_params": null
|
| 141 |
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:846a13b1247bfc212e1559f5a61140830e2e39df0b02326620e4db782b0aaf0c
|
| 3 |
+
size 4792
|