Upload folder using huggingface_hub
Browse files
checkpoint-latest/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 62293080
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:05526172b6b183727ed691244d8fca12632b8705fa3b5f5582a369439119e9f7
|
| 3 |
size 62293080
|
checkpoint-latest/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 124642443
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3ae74e3c93a2f232f2571925d5902330db686e4ed3351b82bafa9a5dbb3dca60
|
| 3 |
size 124642443
|
checkpoint-latest/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:111dc925ae2f7248f73dcd9582f2230af09b95cb292fde39121b77d0a24595e6
|
| 3 |
size 14645
|
checkpoint-latest/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ec9761731a5aa3d5575ffed14f4165db19c93fbedd2eb3e25069def656bd6d84
|
| 3 |
size 1465
|
checkpoint-latest/trainer_state.json
CHANGED
|
@@ -2,7 +2,7 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 500,
|
| 7 |
"global_step": 1000,
|
| 8 |
"is_hyper_param_search": false,
|
|
@@ -10,52 +10,52 @@
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"epoch": 0.
|
| 14 |
-
"grad_norm":
|
| 15 |
-
"learning_rate":
|
| 16 |
-
"loss":
|
| 17 |
-
"step":
|
| 18 |
},
|
| 19 |
{
|
| 20 |
-
"epoch": 0.
|
| 21 |
-
"grad_norm":
|
| 22 |
-
"learning_rate":
|
| 23 |
-
"loss":
|
| 24 |
-
"step":
|
| 25 |
},
|
| 26 |
{
|
| 27 |
-
"epoch": 0.
|
| 28 |
-
"eval_loss":
|
| 29 |
-
"eval_runtime":
|
| 30 |
-
"eval_samples_per_second":
|
| 31 |
-
"eval_steps_per_second":
|
| 32 |
"step": 500
|
| 33 |
},
|
| 34 |
{
|
| 35 |
-
"epoch": 0.
|
| 36 |
-
"grad_norm":
|
| 37 |
-
"learning_rate":
|
| 38 |
-
"loss":
|
| 39 |
-
"step":
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
-
"grad_norm":
|
| 44 |
-
"learning_rate":
|
| 45 |
-
"loss":
|
| 46 |
-
"step":
|
| 47 |
},
|
| 48 |
{
|
| 49 |
-
"epoch": 0.
|
| 50 |
-
"eval_loss":
|
| 51 |
-
"eval_runtime":
|
| 52 |
-
"eval_samples_per_second":
|
| 53 |
-
"eval_steps_per_second":
|
| 54 |
"step": 1000
|
| 55 |
}
|
| 56 |
],
|
| 57 |
-
"logging_steps":
|
| 58 |
-
"max_steps":
|
| 59 |
"num_input_tokens_seen": 0,
|
| 60 |
"num_train_epochs": 3,
|
| 61 |
"save_steps": 1000,
|
|
@@ -71,8 +71,8 @@
|
|
| 71 |
"attributes": {}
|
| 72 |
}
|
| 73 |
},
|
| 74 |
-
"total_flos":
|
| 75 |
-
"train_batch_size":
|
| 76 |
"trial_name": null,
|
| 77 |
"trial_params": null
|
| 78 |
}
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.10828370330265295,
|
| 6 |
"eval_steps": 500,
|
| 7 |
"global_step": 1000,
|
| 8 |
"is_hyper_param_search": false,
|
|
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"epoch": 0.02707092582566324,
|
| 14 |
+
"grad_norm": 0.8431992530822754,
|
| 15 |
+
"learning_rate": 9.910124526258799e-05,
|
| 16 |
+
"loss": 4.3576,
|
| 17 |
+
"step": 250
|
| 18 |
},
|
| 19 |
{
|
| 20 |
+
"epoch": 0.05414185165132648,
|
| 21 |
+
"grad_norm": 0.8061181306838989,
|
| 22 |
+
"learning_rate": 9.819888106839921e-05,
|
| 23 |
+
"loss": 4.2775,
|
| 24 |
+
"step": 500
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"epoch": 0.05414185165132648,
|
| 28 |
+
"eval_loss": 3.5462777614593506,
|
| 29 |
+
"eval_runtime": 87.734,
|
| 30 |
+
"eval_samples_per_second": 112.271,
|
| 31 |
+
"eval_steps_per_second": 7.021,
|
| 32 |
"step": 500
|
| 33 |
},
|
| 34 |
{
|
| 35 |
+
"epoch": 0.08121277747698971,
|
| 36 |
+
"grad_norm": 0.7875335812568665,
|
| 37 |
+
"learning_rate": 9.729651687421044e-05,
|
| 38 |
+
"loss": 4.1927,
|
| 39 |
+
"step": 750
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.10828370330265295,
|
| 43 |
+
"grad_norm": 1.0330173969268799,
|
| 44 |
+
"learning_rate": 9.639415268002166e-05,
|
| 45 |
+
"loss": 4.1178,
|
| 46 |
+
"step": 1000
|
| 47 |
},
|
| 48 |
{
|
| 49 |
+
"epoch": 0.10828370330265295,
|
| 50 |
+
"eval_loss": 3.3765828609466553,
|
| 51 |
+
"eval_runtime": 87.6227,
|
| 52 |
+
"eval_samples_per_second": 112.414,
|
| 53 |
+
"eval_steps_per_second": 7.03,
|
| 54 |
"step": 1000
|
| 55 |
}
|
| 56 |
],
|
| 57 |
+
"logging_steps": 250,
|
| 58 |
+
"max_steps": 27705,
|
| 59 |
"num_input_tokens_seen": 0,
|
| 60 |
"num_train_epochs": 3,
|
| 61 |
"save_steps": 1000,
|
|
|
|
| 71 |
"attributes": {}
|
| 72 |
}
|
| 73 |
},
|
| 74 |
+
"total_flos": 361054076928000.0,
|
| 75 |
+
"train_batch_size": 16,
|
| 76 |
"trial_name": null,
|
| 77 |
"trial_params": null
|
| 78 |
}
|
checkpoint-latest/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6033
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f217a06d41692f3ee4559764c88c0b3c805db6ef1541a53dc2503e0626aeaf07
|
| 3 |
size 6033
|