Upload folder using huggingface_hub
Browse files- model-00001-of-00003.safetensors +1 -1
- model-00002-of-00003.safetensors +1 -1
- model-00003-of-00003.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +41 -3
model-00001-of-00003.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4991044576
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2296dbb3a5d79870e31a3e1d66eef858da04d4331dfa635cbb7466ddb18c8f4
|
| 3 |
size 4991044576
|
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:bf749502b196236c70f506372452fc7a5eeed5d2bfaeb1848b7d63229a53e3bf
|
| 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 4588398592
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5d576f3631a4bb7fa8aa80e433004b0a8e7c68fda5951b66149741056eff6fe4
|
| 3 |
size 4588398592
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 29158711534
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:418a2baa2b53aec26e5d718d3d5a0420ad2a6e360a2fbffba92ce5c96bc6c836
|
| 3 |
size 29158711534
|
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:c3fcb8b7132fdda989f7bbb14a5bf464435849629fe731ccbc64c4724068a57e
|
| 3 |
size 14244
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1000
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a0b6d6f6df2ab28e8148f87f30a2c6734cda26ecb8b2af4ca8d623424f5d1c46
|
| 3 |
size 1000
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 93,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -127,6 +127,44 @@
|
|
| 127 |
"eval_validation_altalexprivacy_samples_per_second": 1.357,
|
| 128 |
"eval_validation_altalexprivacy_steps_per_second": 0.679,
|
| 129 |
"step": 279
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 130 |
}
|
| 131 |
],
|
| 132 |
"logging_steps": 93,
|
|
@@ -134,7 +172,7 @@
|
|
| 134 |
"num_input_tokens_seen": 0,
|
| 135 |
"num_train_epochs": 2,
|
| 136 |
"save_steps": 93,
|
| 137 |
-
"total_flos":
|
| 138 |
"train_batch_size": 1,
|
| 139 |
"trial_name": null,
|
| 140 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.310210038046594,
|
| 5 |
"eval_steps": 93,
|
| 6 |
+
"global_step": 372,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 127 |
"eval_validation_altalexprivacy_samples_per_second": 1.357,
|
| 128 |
"eval_validation_altalexprivacy_steps_per_second": 0.679,
|
| 129 |
"step": 279
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.31,
|
| 133 |
+
"learning_rate": 1e-05,
|
| 134 |
+
"loss": 1.5249,
|
| 135 |
+
"step": 372
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.31,
|
| 139 |
+
"eval_validation_privacy_sources_loss": 1.1636546850204468,
|
| 140 |
+
"eval_validation_privacy_sources_runtime": 169.3483,
|
| 141 |
+
"eval_validation_privacy_sources_samples_per_second": 1.358,
|
| 142 |
+
"eval_validation_privacy_sources_steps_per_second": 0.679,
|
| 143 |
+
"step": 372
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.31,
|
| 147 |
+
"eval_validation_agenda_digitale_loss": 1.839213490486145,
|
| 148 |
+
"eval_validation_agenda_digitale_runtime": 160.5345,
|
| 149 |
+
"eval_validation_agenda_digitale_samples_per_second": 1.358,
|
| 150 |
+
"eval_validation_agenda_digitale_steps_per_second": 0.679,
|
| 151 |
+
"step": 372
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 0.31,
|
| 155 |
+
"eval_validation_leggepertutti_loss": 1.6400021314620972,
|
| 156 |
+
"eval_validation_leggepertutti_runtime": 27.9724,
|
| 157 |
+
"eval_validation_leggepertutti_samples_per_second": 1.358,
|
| 158 |
+
"eval_validation_leggepertutti_steps_per_second": 0.679,
|
| 159 |
+
"step": 372
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.31,
|
| 163 |
+
"eval_validation_altalexprivacy_loss": 1.6672061681747437,
|
| 164 |
+
"eval_validation_altalexprivacy_runtime": 45.6542,
|
| 165 |
+
"eval_validation_altalexprivacy_samples_per_second": 1.358,
|
| 166 |
+
"eval_validation_altalexprivacy_steps_per_second": 0.679,
|
| 167 |
+
"step": 372
|
| 168 |
}
|
| 169 |
],
|
| 170 |
"logging_steps": 93,
|
|
|
|
| 172 |
"num_input_tokens_seen": 0,
|
| 173 |
"num_train_epochs": 2,
|
| 174 |
"save_steps": 93,
|
| 175 |
+
"total_flos": 4.174493755404976e+18,
|
| 176 |
"train_batch_size": 1,
|
| 177 |
"trial_name": null,
|
| 178 |
"trial_params": null
|