wangrongsheng
commited on
Commit
·
dbd399b
1
Parent(s):
ef83107
commit from root
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- README.md +3 -0
- adapter_config.json +19 -0
- adapter_model.bin +3 -0
- all_results.json +7 -0
- checkpoint-1000/README.md +3 -0
- checkpoint-1000/adapter_config.json +19 -0
- checkpoint-1000/adapter_model.bin +3 -0
- checkpoint-1000/finetuning_args.json +13 -0
- checkpoint-1000/optimizer.pt +3 -0
- checkpoint-1000/rng_state_0.pth +3 -0
- checkpoint-1000/rng_state_1.pth +3 -0
- checkpoint-1000/rng_state_2.pth +3 -0
- checkpoint-1000/rng_state_3.pth +3 -0
- checkpoint-1000/scheduler.pt +3 -0
- checkpoint-1000/trainer_state.json +616 -0
- checkpoint-1000/training_args.bin +3 -0
- checkpoint-10000/README.md +3 -0
- checkpoint-10000/adapter_config.json +19 -0
- checkpoint-10000/adapter_model.bin +3 -0
- checkpoint-10000/finetuning_args.json +13 -0
- checkpoint-10000/optimizer.pt +3 -0
- checkpoint-10000/rng_state_0.pth +3 -0
- checkpoint-10000/rng_state_1.pth +3 -0
- checkpoint-10000/rng_state_2.pth +3 -0
- checkpoint-10000/rng_state_3.pth +3 -0
- checkpoint-10000/scheduler.pt +3 -0
- checkpoint-10000/trainer_state.json +0 -0
- checkpoint-10000/training_args.bin +3 -0
- checkpoint-2000/README.md +3 -0
- checkpoint-2000/adapter_config.json +19 -0
- checkpoint-2000/adapter_model.bin +3 -0
- checkpoint-2000/finetuning_args.json +13 -0
- checkpoint-2000/optimizer.pt +3 -0
- checkpoint-2000/rng_state_0.pth +3 -0
- checkpoint-2000/rng_state_1.pth +3 -0
- checkpoint-2000/rng_state_2.pth +3 -0
- checkpoint-2000/rng_state_3.pth +3 -0
- checkpoint-2000/scheduler.pt +3 -0
- checkpoint-2000/trainer_state.json +1216 -0
- checkpoint-2000/training_args.bin +3 -0
- checkpoint-3000/README.md +3 -0
- checkpoint-3000/adapter_config.json +19 -0
- checkpoint-3000/adapter_model.bin +3 -0
- checkpoint-3000/finetuning_args.json +13 -0
- checkpoint-3000/optimizer.pt +3 -0
- checkpoint-3000/rng_state_0.pth +3 -0
- checkpoint-3000/rng_state_1.pth +3 -0
- checkpoint-3000/rng_state_2.pth +3 -0
- checkpoint-3000/rng_state_3.pth +3 -0
- checkpoint-3000/scheduler.pt +3 -0
README.md
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
adapter_config.json
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"base_model_name_or_path": "chatglm2-6b",
|
| 3 |
+
"bias": "none",
|
| 4 |
+
"fan_in_fan_out": false,
|
| 5 |
+
"inference_mode": true,
|
| 6 |
+
"init_lora_weights": true,
|
| 7 |
+
"layers_pattern": null,
|
| 8 |
+
"layers_to_transform": null,
|
| 9 |
+
"lora_alpha": 32.0,
|
| 10 |
+
"lora_dropout": 0.1,
|
| 11 |
+
"modules_to_save": null,
|
| 12 |
+
"peft_type": "LORA",
|
| 13 |
+
"r": 8,
|
| 14 |
+
"revision": null,
|
| 15 |
+
"target_modules": [
|
| 16 |
+
"query_key_value"
|
| 17 |
+
],
|
| 18 |
+
"task_type": "CAUSAL_LM"
|
| 19 |
+
}
|
adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c3c6758d97d365c10604d10e27b6be873c3ad2e83a9785463a6ba10bfe630d8e
|
| 3 |
+
size 7819417
|
all_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"train_loss": 1.978195481530868,
|
| 4 |
+
"train_runtime": 26343.521,
|
| 5 |
+
"train_samples_per_second": 26.232,
|
| 6 |
+
"train_steps_per_second": 0.41
|
| 7 |
+
}
|
checkpoint-1000/README.md
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
checkpoint-1000/adapter_config.json
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"base_model_name_or_path": "chatglm2-6b",
|
| 3 |
+
"bias": "none",
|
| 4 |
+
"fan_in_fan_out": false,
|
| 5 |
+
"inference_mode": true,
|
| 6 |
+
"init_lora_weights": true,
|
| 7 |
+
"layers_pattern": null,
|
| 8 |
+
"layers_to_transform": null,
|
| 9 |
+
"lora_alpha": 32.0,
|
| 10 |
+
"lora_dropout": 0.1,
|
| 11 |
+
"modules_to_save": null,
|
| 12 |
+
"peft_type": "LORA",
|
| 13 |
+
"r": 8,
|
| 14 |
+
"revision": null,
|
| 15 |
+
"target_modules": [
|
| 16 |
+
"query_key_value"
|
| 17 |
+
],
|
| 18 |
+
"task_type": "CAUSAL_LM"
|
| 19 |
+
}
|
checkpoint-1000/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d168ef0b419dc788a731c7120dada236c6fbae4d8e809d70c45f93f345fa67ab
|
| 3 |
+
size 7819417
|
checkpoint-1000/finetuning_args.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"finetuning_type": "lora",
|
| 3 |
+
"lora_alpha": 32.0,
|
| 4 |
+
"lora_dropout": 0.1,
|
| 5 |
+
"lora_rank": 8,
|
| 6 |
+
"lora_target": [
|
| 7 |
+
"query_key_value"
|
| 8 |
+
],
|
| 9 |
+
"name_module_trainable": "mlp",
|
| 10 |
+
"num_layer_trainable": 3,
|
| 11 |
+
"pre_seq_len": 64,
|
| 12 |
+
"prefix_projection": false
|
| 13 |
+
}
|
checkpoint-1000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:207da355524f4452d88bdbb9d13f0ed7b4aa02f827dc0adeaa53dd3c4a1d13a1
|
| 3 |
+
size 15644485
|
checkpoint-1000/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0a2684ce62e8d7c210c23c57d059a952c18a26260da209f1d6ebe314693e4974
|
| 3 |
+
size 18679
|
checkpoint-1000/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:de719cf6845a3b19b96284a78513c960a388c70f13766d1ea991700b32e79d53
|
| 3 |
+
size 18679
|
checkpoint-1000/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:58e0772c04da6662579787b6678d4d2795a96e4da23118d565f8dd6c2e25617b
|
| 3 |
+
size 18679
|
checkpoint-1000/rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:55f40b16d80d11c0cf20f8dc79b69affdfca686fdae7995161a0db1c59c9ceed
|
| 3 |
+
size 18679
|
checkpoint-1000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6f01619e7f8d789aa01ef80c209d2e88c5d1080670973e05d35c732aa495fa39
|
| 3 |
+
size 627
|
checkpoint-1000/trainer_state.json
ADDED
|
@@ -0,0 +1,616 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.18521948508983144,
|
| 5 |
+
"global_step": 1000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
+
"learning_rate": 0.0009999978838190456,
|
| 13 |
+
"loss": 2.9794,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.0,
|
| 18 |
+
"learning_rate": 0.0009999915352940948,
|
| 19 |
+
"loss": 2.3885,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 0.000999980954478887,
|
| 25 |
+
"loss": 2.3057,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.01,
|
| 30 |
+
"learning_rate": 0.000999966141462985,
|
| 31 |
+
"loss": 2.2692,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.01,
|
| 36 |
+
"learning_rate": 0.000999947096371777,
|
| 37 |
+
"loss": 2.2576,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.01,
|
| 42 |
+
"learning_rate": 0.0009999238193664748,
|
| 43 |
+
"loss": 2.2388,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.01,
|
| 48 |
+
"learning_rate": 0.0009998963106441117,
|
| 49 |
+
"loss": 2.2523,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.01,
|
| 54 |
+
"learning_rate": 0.0009998645704375414,
|
| 55 |
+
"loss": 2.218,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.02,
|
| 60 |
+
"learning_rate": 0.000999828599015436,
|
| 61 |
+
"loss": 2.2457,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.02,
|
| 66 |
+
"learning_rate": 0.0009997883966822835,
|
| 67 |
+
"loss": 2.198,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.02,
|
| 72 |
+
"learning_rate": 0.0009997439637783859,
|
| 73 |
+
"loss": 2.2013,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.02,
|
| 78 |
+
"learning_rate": 0.000999695300679855,
|
| 79 |
+
"loss": 2.1765,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.02,
|
| 84 |
+
"learning_rate": 0.0009996424077986109,
|
| 85 |
+
"loss": 2.1741,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.03,
|
| 90 |
+
"learning_rate": 0.000999585285582377,
|
| 91 |
+
"loss": 2.1898,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.03,
|
| 96 |
+
"learning_rate": 0.0009995239345146772,
|
| 97 |
+
"loss": 2.1466,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.03,
|
| 102 |
+
"learning_rate": 0.0009994583551148314,
|
| 103 |
+
"loss": 2.1423,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.03,
|
| 108 |
+
"learning_rate": 0.0009993885479379506,
|
| 109 |
+
"loss": 2.1451,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.03,
|
| 114 |
+
"learning_rate": 0.000999314513574934,
|
| 115 |
+
"loss": 2.202,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.04,
|
| 120 |
+
"learning_rate": 0.0009992362526524616,
|
| 121 |
+
"loss": 2.1208,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.04,
|
| 126 |
+
"learning_rate": 0.0009991537658329906,
|
| 127 |
+
"loss": 2.1591,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.04,
|
| 132 |
+
"learning_rate": 0.000999067053814749,
|
| 133 |
+
"loss": 2.1788,
|
| 134 |
+
"step": 210
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.04,
|
| 138 |
+
"learning_rate": 0.0009989761173317304,
|
| 139 |
+
"loss": 2.147,
|
| 140 |
+
"step": 220
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.04,
|
| 144 |
+
"learning_rate": 0.000998880957153687,
|
| 145 |
+
"loss": 2.1249,
|
| 146 |
+
"step": 230
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.04,
|
| 150 |
+
"learning_rate": 0.000998781574086123,
|
| 151 |
+
"loss": 2.165,
|
| 152 |
+
"step": 240
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.05,
|
| 156 |
+
"learning_rate": 0.000998677968970289,
|
| 157 |
+
"loss": 2.1428,
|
| 158 |
+
"step": 250
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.05,
|
| 162 |
+
"learning_rate": 0.0009985701426831735,
|
| 163 |
+
"loss": 2.1384,
|
| 164 |
+
"step": 260
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.05,
|
| 168 |
+
"learning_rate": 0.0009984580961374964,
|
| 169 |
+
"loss": 2.1585,
|
| 170 |
+
"step": 270
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.05,
|
| 174 |
+
"learning_rate": 0.0009983418302817008,
|
| 175 |
+
"loss": 2.1156,
|
| 176 |
+
"step": 280
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.05,
|
| 180 |
+
"learning_rate": 0.0009982213460999448,
|
| 181 |
+
"loss": 2.0811,
|
| 182 |
+
"step": 290
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.06,
|
| 186 |
+
"learning_rate": 0.000998096644612094,
|
| 187 |
+
"loss": 2.1081,
|
| 188 |
+
"step": 300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.06,
|
| 192 |
+
"learning_rate": 0.0009979677268737118,
|
| 193 |
+
"loss": 2.1246,
|
| 194 |
+
"step": 310
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.06,
|
| 198 |
+
"learning_rate": 0.0009978345939760515,
|
| 199 |
+
"loss": 2.1229,
|
| 200 |
+
"step": 320
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.06,
|
| 204 |
+
"learning_rate": 0.000997697247046046,
|
| 205 |
+
"loss": 2.1033,
|
| 206 |
+
"step": 330
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.06,
|
| 210 |
+
"learning_rate": 0.0009975556872462994,
|
| 211 |
+
"loss": 2.0931,
|
| 212 |
+
"step": 340
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.06,
|
| 216 |
+
"learning_rate": 0.000997409915775076,
|
| 217 |
+
"loss": 2.1206,
|
| 218 |
+
"step": 350
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.07,
|
| 222 |
+
"learning_rate": 0.0009972599338662915,
|
| 223 |
+
"loss": 2.0537,
|
| 224 |
+
"step": 360
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.07,
|
| 228 |
+
"learning_rate": 0.0009971057427895012,
|
| 229 |
+
"loss": 2.0762,
|
| 230 |
+
"step": 370
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.07,
|
| 234 |
+
"learning_rate": 0.0009969473438498897,
|
| 235 |
+
"loss": 2.0883,
|
| 236 |
+
"step": 380
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.07,
|
| 240 |
+
"learning_rate": 0.0009967847383882604,
|
| 241 |
+
"loss": 2.1174,
|
| 242 |
+
"step": 390
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.07,
|
| 246 |
+
"learning_rate": 0.0009966179277810239,
|
| 247 |
+
"loss": 2.1111,
|
| 248 |
+
"step": 400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.08,
|
| 252 |
+
"learning_rate": 0.0009964469134401855,
|
| 253 |
+
"loss": 2.1288,
|
| 254 |
+
"step": 410
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.08,
|
| 258 |
+
"learning_rate": 0.0009962716968133346,
|
| 259 |
+
"loss": 2.0967,
|
| 260 |
+
"step": 420
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.08,
|
| 264 |
+
"learning_rate": 0.0009960922793836318,
|
| 265 |
+
"loss": 2.1216,
|
| 266 |
+
"step": 430
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.08,
|
| 270 |
+
"learning_rate": 0.0009959086626697955,
|
| 271 |
+
"loss": 2.0924,
|
| 272 |
+
"step": 440
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.08,
|
| 276 |
+
"learning_rate": 0.0009957208482260908,
|
| 277 |
+
"loss": 2.0809,
|
| 278 |
+
"step": 450
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.09,
|
| 282 |
+
"learning_rate": 0.0009955288376423152,
|
| 283 |
+
"loss": 2.1082,
|
| 284 |
+
"step": 460
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.09,
|
| 288 |
+
"learning_rate": 0.0009953326325437852,
|
| 289 |
+
"loss": 2.0885,
|
| 290 |
+
"step": 470
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.09,
|
| 294 |
+
"learning_rate": 0.0009951322345913224,
|
| 295 |
+
"loss": 2.1133,
|
| 296 |
+
"step": 480
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.09,
|
| 300 |
+
"learning_rate": 0.0009949276454812408,
|
| 301 |
+
"loss": 2.0844,
|
| 302 |
+
"step": 490
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.09,
|
| 306 |
+
"learning_rate": 0.00099471886694533,
|
| 307 |
+
"loss": 2.0796,
|
| 308 |
+
"step": 500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.09,
|
| 312 |
+
"learning_rate": 0.0009945059007508434,
|
| 313 |
+
"loss": 2.1255,
|
| 314 |
+
"step": 510
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.1,
|
| 318 |
+
"learning_rate": 0.0009942887487004804,
|
| 319 |
+
"loss": 2.0913,
|
| 320 |
+
"step": 520
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.1,
|
| 324 |
+
"learning_rate": 0.0009940674126323733,
|
| 325 |
+
"loss": 2.1003,
|
| 326 |
+
"step": 530
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.1,
|
| 330 |
+
"learning_rate": 0.0009938418944200709,
|
| 331 |
+
"loss": 2.0541,
|
| 332 |
+
"step": 540
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.1,
|
| 336 |
+
"learning_rate": 0.0009936121959725223,
|
| 337 |
+
"loss": 2.0523,
|
| 338 |
+
"step": 550
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.1,
|
| 342 |
+
"learning_rate": 0.0009933783192340618,
|
| 343 |
+
"loss": 2.1225,
|
| 344 |
+
"step": 560
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.11,
|
| 348 |
+
"learning_rate": 0.0009931402661843911,
|
| 349 |
+
"loss": 2.0446,
|
| 350 |
+
"step": 570
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.11,
|
| 354 |
+
"learning_rate": 0.000992898038838564,
|
| 355 |
+
"loss": 2.0921,
|
| 356 |
+
"step": 580
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.11,
|
| 360 |
+
"learning_rate": 0.0009926516392469674,
|
| 361 |
+
"loss": 2.1081,
|
| 362 |
+
"step": 590
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.11,
|
| 366 |
+
"learning_rate": 0.0009924010694953064,
|
| 367 |
+
"loss": 2.0734,
|
| 368 |
+
"step": 600
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.11,
|
| 372 |
+
"learning_rate": 0.0009921463317045843,
|
| 373 |
+
"loss": 2.0652,
|
| 374 |
+
"step": 610
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.11,
|
| 378 |
+
"learning_rate": 0.0009918874280310862,
|
| 379 |
+
"loss": 2.0818,
|
| 380 |
+
"step": 620
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.12,
|
| 384 |
+
"learning_rate": 0.0009916243606663605,
|
| 385 |
+
"loss": 2.0776,
|
| 386 |
+
"step": 630
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.12,
|
| 390 |
+
"learning_rate": 0.0009913571318371994,
|
| 391 |
+
"loss": 2.1025,
|
| 392 |
+
"step": 640
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.12,
|
| 396 |
+
"learning_rate": 0.0009910857438056215,
|
| 397 |
+
"loss": 2.066,
|
| 398 |
+
"step": 650
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.12,
|
| 402 |
+
"learning_rate": 0.0009908101988688512,
|
| 403 |
+
"loss": 2.0575,
|
| 404 |
+
"step": 660
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.12,
|
| 408 |
+
"learning_rate": 0.0009905304993593008,
|
| 409 |
+
"loss": 2.1269,
|
| 410 |
+
"step": 670
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.13,
|
| 414 |
+
"learning_rate": 0.0009902466476445486,
|
| 415 |
+
"loss": 2.0518,
|
| 416 |
+
"step": 680
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.13,
|
| 420 |
+
"learning_rate": 0.0009899586461273218,
|
| 421 |
+
"loss": 2.0698,
|
| 422 |
+
"step": 690
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.13,
|
| 426 |
+
"learning_rate": 0.000989666497245473,
|
| 427 |
+
"loss": 2.0988,
|
| 428 |
+
"step": 700
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.13,
|
| 432 |
+
"learning_rate": 0.0009893702034719624,
|
| 433 |
+
"loss": 2.0986,
|
| 434 |
+
"step": 710
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.13,
|
| 438 |
+
"learning_rate": 0.0009890697673148345,
|
| 439 |
+
"loss": 2.0237,
|
| 440 |
+
"step": 720
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.14,
|
| 444 |
+
"learning_rate": 0.0009887651913171986,
|
| 445 |
+
"loss": 2.0027,
|
| 446 |
+
"step": 730
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.14,
|
| 450 |
+
"learning_rate": 0.0009884564780572064,
|
| 451 |
+
"loss": 2.0563,
|
| 452 |
+
"step": 740
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.14,
|
| 456 |
+
"learning_rate": 0.0009881436301480305,
|
| 457 |
+
"loss": 2.0624,
|
| 458 |
+
"step": 750
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.14,
|
| 462 |
+
"learning_rate": 0.000987826650237842,
|
| 463 |
+
"loss": 2.0926,
|
| 464 |
+
"step": 760
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.14,
|
| 468 |
+
"learning_rate": 0.000987505541009788,
|
| 469 |
+
"loss": 2.0585,
|
| 470 |
+
"step": 770
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.14,
|
| 474 |
+
"learning_rate": 0.0009871803051819696,
|
| 475 |
+
"loss": 2.0494,
|
| 476 |
+
"step": 780
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.15,
|
| 480 |
+
"learning_rate": 0.0009868509455074183,
|
| 481 |
+
"loss": 2.0106,
|
| 482 |
+
"step": 790
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.15,
|
| 486 |
+
"learning_rate": 0.0009865174647740729,
|
| 487 |
+
"loss": 2.0861,
|
| 488 |
+
"step": 800
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.15,
|
| 492 |
+
"learning_rate": 0.0009861798658047556,
|
| 493 |
+
"loss": 2.0478,
|
| 494 |
+
"step": 810
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.15,
|
| 498 |
+
"learning_rate": 0.0009858381514571484,
|
| 499 |
+
"loss": 2.0469,
|
| 500 |
+
"step": 820
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.15,
|
| 504 |
+
"learning_rate": 0.000985492324623769,
|
| 505 |
+
"loss": 2.0671,
|
| 506 |
+
"step": 830
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.16,
|
| 510 |
+
"learning_rate": 0.0009851423882319458,
|
| 511 |
+
"loss": 2.0808,
|
| 512 |
+
"step": 840
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.16,
|
| 516 |
+
"learning_rate": 0.0009847883452437937,
|
| 517 |
+
"loss": 2.0331,
|
| 518 |
+
"step": 850
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.16,
|
| 522 |
+
"learning_rate": 0.0009844301986561893,
|
| 523 |
+
"loss": 2.0295,
|
| 524 |
+
"step": 860
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.16,
|
| 528 |
+
"learning_rate": 0.000984067951500744,
|
| 529 |
+
"loss": 2.0873,
|
| 530 |
+
"step": 870
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.16,
|
| 534 |
+
"learning_rate": 0.00098370160684378,
|
| 535 |
+
"loss": 2.1038,
|
| 536 |
+
"step": 880
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.16,
|
| 540 |
+
"learning_rate": 0.0009833311677863042,
|
| 541 |
+
"loss": 2.0337,
|
| 542 |
+
"step": 890
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.17,
|
| 546 |
+
"learning_rate": 0.0009829566374639801,
|
| 547 |
+
"loss": 2.0407,
|
| 548 |
+
"step": 900
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.17,
|
| 552 |
+
"learning_rate": 0.0009825780190471042,
|
| 553 |
+
"loss": 2.1049,
|
| 554 |
+
"step": 910
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.17,
|
| 558 |
+
"learning_rate": 0.000982195315740576,
|
| 559 |
+
"loss": 2.0475,
|
| 560 |
+
"step": 920
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.17,
|
| 564 |
+
"learning_rate": 0.0009818085307838741,
|
| 565 |
+
"loss": 2.0624,
|
| 566 |
+
"step": 930
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.17,
|
| 570 |
+
"learning_rate": 0.000981417667451026,
|
| 571 |
+
"loss": 2.0714,
|
| 572 |
+
"step": 940
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.18,
|
| 576 |
+
"learning_rate": 0.0009810227290505816,
|
| 577 |
+
"loss": 2.0947,
|
| 578 |
+
"step": 950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.18,
|
| 582 |
+
"learning_rate": 0.0009806237189255859,
|
| 583 |
+
"loss": 2.0591,
|
| 584 |
+
"step": 960
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.18,
|
| 588 |
+
"learning_rate": 0.0009802206404535489,
|
| 589 |
+
"loss": 2.0301,
|
| 590 |
+
"step": 970
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.18,
|
| 594 |
+
"learning_rate": 0.000979813497046419,
|
| 595 |
+
"loss": 2.0556,
|
| 596 |
+
"step": 980
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.18,
|
| 600 |
+
"learning_rate": 0.0009794022921505523,
|
| 601 |
+
"loss": 2.0753,
|
| 602 |
+
"step": 990
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.19,
|
| 606 |
+
"learning_rate": 0.000978987029246685,
|
| 607 |
+
"loss": 2.0898,
|
| 608 |
+
"step": 1000
|
| 609 |
+
}
|
| 610 |
+
],
|
| 611 |
+
"max_steps": 10798,
|
| 612 |
+
"num_train_epochs": 2,
|
| 613 |
+
"total_flos": 9.498667895656284e+17,
|
| 614 |
+
"trial_name": null,
|
| 615 |
+
"trial_params": null
|
| 616 |
+
}
|
checkpoint-1000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8813832b030b0a678cb3ad0abd28db5000bdb8bb1708e45ba10cfda908fcc38
|
| 3 |
+
size 3305
|
checkpoint-10000/README.md
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
checkpoint-10000/adapter_config.json
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"base_model_name_or_path": "chatglm2-6b",
|
| 3 |
+
"bias": "none",
|
| 4 |
+
"fan_in_fan_out": false,
|
| 5 |
+
"inference_mode": true,
|
| 6 |
+
"init_lora_weights": true,
|
| 7 |
+
"layers_pattern": null,
|
| 8 |
+
"layers_to_transform": null,
|
| 9 |
+
"lora_alpha": 32.0,
|
| 10 |
+
"lora_dropout": 0.1,
|
| 11 |
+
"modules_to_save": null,
|
| 12 |
+
"peft_type": "LORA",
|
| 13 |
+
"r": 8,
|
| 14 |
+
"revision": null,
|
| 15 |
+
"target_modules": [
|
| 16 |
+
"query_key_value"
|
| 17 |
+
],
|
| 18 |
+
"task_type": "CAUSAL_LM"
|
| 19 |
+
}
|
checkpoint-10000/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9903abff597676a5d85a3f42ba3f950256d00e4aa93564c8ebb5fd5f4a0ebbbe
|
| 3 |
+
size 7819417
|
checkpoint-10000/finetuning_args.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"finetuning_type": "lora",
|
| 3 |
+
"lora_alpha": 32.0,
|
| 4 |
+
"lora_dropout": 0.1,
|
| 5 |
+
"lora_rank": 8,
|
| 6 |
+
"lora_target": [
|
| 7 |
+
"query_key_value"
|
| 8 |
+
],
|
| 9 |
+
"name_module_trainable": "mlp",
|
| 10 |
+
"num_layer_trainable": 3,
|
| 11 |
+
"pre_seq_len": 64,
|
| 12 |
+
"prefix_projection": false
|
| 13 |
+
}
|
checkpoint-10000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9986fb97ca3fa48e2015b70aa583802fa73acc98d578bb5ac096548c91c4f424
|
| 3 |
+
size 15644485
|
checkpoint-10000/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:37fd716397629e3be922b426ba94f12d00c46352e66dbe11b85822a76d3e35e1
|
| 3 |
+
size 18679
|
checkpoint-10000/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:74bfb7ea10ec3e65f8eca160d3bf18eef130d179b66b4138138d46fceab4daca
|
| 3 |
+
size 18679
|
checkpoint-10000/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:488534d9a95b06aeefe2b45dfed265600bde02bf26605bfacb8ea0bca14920a4
|
| 3 |
+
size 18679
|
checkpoint-10000/rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9edad9c0f7873d58b81e886e6d846f91293545f6932489450d12b86cf6e02700
|
| 3 |
+
size 18679
|
checkpoint-10000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d31ab5731d5270c6c302ae9dc8c05030cf9f832744fe07067f6bed6ea7d85df0
|
| 3 |
+
size 627
|
checkpoint-10000/trainer_state.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-10000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8813832b030b0a678cb3ad0abd28db5000bdb8bb1708e45ba10cfda908fcc38
|
| 3 |
+
size 3305
|
checkpoint-2000/README.md
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
checkpoint-2000/adapter_config.json
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"base_model_name_or_path": "chatglm2-6b",
|
| 3 |
+
"bias": "none",
|
| 4 |
+
"fan_in_fan_out": false,
|
| 5 |
+
"inference_mode": true,
|
| 6 |
+
"init_lora_weights": true,
|
| 7 |
+
"layers_pattern": null,
|
| 8 |
+
"layers_to_transform": null,
|
| 9 |
+
"lora_alpha": 32.0,
|
| 10 |
+
"lora_dropout": 0.1,
|
| 11 |
+
"modules_to_save": null,
|
| 12 |
+
"peft_type": "LORA",
|
| 13 |
+
"r": 8,
|
| 14 |
+
"revision": null,
|
| 15 |
+
"target_modules": [
|
| 16 |
+
"query_key_value"
|
| 17 |
+
],
|
| 18 |
+
"task_type": "CAUSAL_LM"
|
| 19 |
+
}
|
checkpoint-2000/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6f081bbdfbb5e387ee0824fff7f0f47949ffa6e68e7943ffa59134934c4cc853
|
| 3 |
+
size 7819417
|
checkpoint-2000/finetuning_args.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"finetuning_type": "lora",
|
| 3 |
+
"lora_alpha": 32.0,
|
| 4 |
+
"lora_dropout": 0.1,
|
| 5 |
+
"lora_rank": 8,
|
| 6 |
+
"lora_target": [
|
| 7 |
+
"query_key_value"
|
| 8 |
+
],
|
| 9 |
+
"name_module_trainable": "mlp",
|
| 10 |
+
"num_layer_trainable": 3,
|
| 11 |
+
"pre_seq_len": 64,
|
| 12 |
+
"prefix_projection": false
|
| 13 |
+
}
|
checkpoint-2000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3f9fe37c01b31ebef944ca17d10d8d5aaacb5798001799c604caf3846715c32d
|
| 3 |
+
size 15644485
|
checkpoint-2000/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d8cb970e4ced27f56008fda1204003f4d92bee7bbcb658b1d95feeb1ee2d9c9
|
| 3 |
+
size 18679
|
checkpoint-2000/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dc409f2fff3209bb2720c94bd3a28d16c2aea229945134be8dfd24bd233b555a
|
| 3 |
+
size 18679
|
checkpoint-2000/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:994276963fc670cf27a5b4003c9939fb3694b73e3468dcf68c44d82388fb664e
|
| 3 |
+
size 18679
|
checkpoint-2000/rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:19db02794e617e7615491a559802b701bc18a3aebaf0b383447ebdfcb7a6c4ca
|
| 3 |
+
size 18679
|
checkpoint-2000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:74b5152c9a31debe910da86a00a4ef326dfa0b3a55c019894c4686f1176be2b3
|
| 3 |
+
size 627
|
checkpoint-2000/trainer_state.json
ADDED
|
@@ -0,0 +1,1216 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.3704389701796629,
|
| 5 |
+
"global_step": 2000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
+
"learning_rate": 0.0009999978838190456,
|
| 13 |
+
"loss": 2.9794,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.0,
|
| 18 |
+
"learning_rate": 0.0009999915352940948,
|
| 19 |
+
"loss": 2.3885,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 0.000999980954478887,
|
| 25 |
+
"loss": 2.3057,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.01,
|
| 30 |
+
"learning_rate": 0.000999966141462985,
|
| 31 |
+
"loss": 2.2692,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.01,
|
| 36 |
+
"learning_rate": 0.000999947096371777,
|
| 37 |
+
"loss": 2.2576,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.01,
|
| 42 |
+
"learning_rate": 0.0009999238193664748,
|
| 43 |
+
"loss": 2.2388,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.01,
|
| 48 |
+
"learning_rate": 0.0009998963106441117,
|
| 49 |
+
"loss": 2.2523,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.01,
|
| 54 |
+
"learning_rate": 0.0009998645704375414,
|
| 55 |
+
"loss": 2.218,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.02,
|
| 60 |
+
"learning_rate": 0.000999828599015436,
|
| 61 |
+
"loss": 2.2457,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.02,
|
| 66 |
+
"learning_rate": 0.0009997883966822835,
|
| 67 |
+
"loss": 2.198,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.02,
|
| 72 |
+
"learning_rate": 0.0009997439637783859,
|
| 73 |
+
"loss": 2.2013,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.02,
|
| 78 |
+
"learning_rate": 0.000999695300679855,
|
| 79 |
+
"loss": 2.1765,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.02,
|
| 84 |
+
"learning_rate": 0.0009996424077986109,
|
| 85 |
+
"loss": 2.1741,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.03,
|
| 90 |
+
"learning_rate": 0.000999585285582377,
|
| 91 |
+
"loss": 2.1898,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.03,
|
| 96 |
+
"learning_rate": 0.0009995239345146772,
|
| 97 |
+
"loss": 2.1466,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.03,
|
| 102 |
+
"learning_rate": 0.0009994583551148314,
|
| 103 |
+
"loss": 2.1423,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.03,
|
| 108 |
+
"learning_rate": 0.0009993885479379506,
|
| 109 |
+
"loss": 2.1451,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.03,
|
| 114 |
+
"learning_rate": 0.000999314513574934,
|
| 115 |
+
"loss": 2.202,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.04,
|
| 120 |
+
"learning_rate": 0.0009992362526524616,
|
| 121 |
+
"loss": 2.1208,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.04,
|
| 126 |
+
"learning_rate": 0.0009991537658329906,
|
| 127 |
+
"loss": 2.1591,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.04,
|
| 132 |
+
"learning_rate": 0.000999067053814749,
|
| 133 |
+
"loss": 2.1788,
|
| 134 |
+
"step": 210
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.04,
|
| 138 |
+
"learning_rate": 0.0009989761173317304,
|
| 139 |
+
"loss": 2.147,
|
| 140 |
+
"step": 220
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.04,
|
| 144 |
+
"learning_rate": 0.000998880957153687,
|
| 145 |
+
"loss": 2.1249,
|
| 146 |
+
"step": 230
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.04,
|
| 150 |
+
"learning_rate": 0.000998781574086123,
|
| 151 |
+
"loss": 2.165,
|
| 152 |
+
"step": 240
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.05,
|
| 156 |
+
"learning_rate": 0.000998677968970289,
|
| 157 |
+
"loss": 2.1428,
|
| 158 |
+
"step": 250
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.05,
|
| 162 |
+
"learning_rate": 0.0009985701426831735,
|
| 163 |
+
"loss": 2.1384,
|
| 164 |
+
"step": 260
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.05,
|
| 168 |
+
"learning_rate": 0.0009984580961374964,
|
| 169 |
+
"loss": 2.1585,
|
| 170 |
+
"step": 270
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.05,
|
| 174 |
+
"learning_rate": 0.0009983418302817008,
|
| 175 |
+
"loss": 2.1156,
|
| 176 |
+
"step": 280
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.05,
|
| 180 |
+
"learning_rate": 0.0009982213460999448,
|
| 181 |
+
"loss": 2.0811,
|
| 182 |
+
"step": 290
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.06,
|
| 186 |
+
"learning_rate": 0.000998096644612094,
|
| 187 |
+
"loss": 2.1081,
|
| 188 |
+
"step": 300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.06,
|
| 192 |
+
"learning_rate": 0.0009979677268737118,
|
| 193 |
+
"loss": 2.1246,
|
| 194 |
+
"step": 310
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.06,
|
| 198 |
+
"learning_rate": 0.0009978345939760515,
|
| 199 |
+
"loss": 2.1229,
|
| 200 |
+
"step": 320
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.06,
|
| 204 |
+
"learning_rate": 0.000997697247046046,
|
| 205 |
+
"loss": 2.1033,
|
| 206 |
+
"step": 330
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.06,
|
| 210 |
+
"learning_rate": 0.0009975556872462994,
|
| 211 |
+
"loss": 2.0931,
|
| 212 |
+
"step": 340
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.06,
|
| 216 |
+
"learning_rate": 0.000997409915775076,
|
| 217 |
+
"loss": 2.1206,
|
| 218 |
+
"step": 350
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.07,
|
| 222 |
+
"learning_rate": 0.0009972599338662915,
|
| 223 |
+
"loss": 2.0537,
|
| 224 |
+
"step": 360
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.07,
|
| 228 |
+
"learning_rate": 0.0009971057427895012,
|
| 229 |
+
"loss": 2.0762,
|
| 230 |
+
"step": 370
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.07,
|
| 234 |
+
"learning_rate": 0.0009969473438498897,
|
| 235 |
+
"loss": 2.0883,
|
| 236 |
+
"step": 380
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.07,
|
| 240 |
+
"learning_rate": 0.0009967847383882604,
|
| 241 |
+
"loss": 2.1174,
|
| 242 |
+
"step": 390
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.07,
|
| 246 |
+
"learning_rate": 0.0009966179277810239,
|
| 247 |
+
"loss": 2.1111,
|
| 248 |
+
"step": 400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.08,
|
| 252 |
+
"learning_rate": 0.0009964469134401855,
|
| 253 |
+
"loss": 2.1288,
|
| 254 |
+
"step": 410
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.08,
|
| 258 |
+
"learning_rate": 0.0009962716968133346,
|
| 259 |
+
"loss": 2.0967,
|
| 260 |
+
"step": 420
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.08,
|
| 264 |
+
"learning_rate": 0.0009960922793836318,
|
| 265 |
+
"loss": 2.1216,
|
| 266 |
+
"step": 430
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.08,
|
| 270 |
+
"learning_rate": 0.0009959086626697955,
|
| 271 |
+
"loss": 2.0924,
|
| 272 |
+
"step": 440
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.08,
|
| 276 |
+
"learning_rate": 0.0009957208482260908,
|
| 277 |
+
"loss": 2.0809,
|
| 278 |
+
"step": 450
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.09,
|
| 282 |
+
"learning_rate": 0.0009955288376423152,
|
| 283 |
+
"loss": 2.1082,
|
| 284 |
+
"step": 460
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.09,
|
| 288 |
+
"learning_rate": 0.0009953326325437852,
|
| 289 |
+
"loss": 2.0885,
|
| 290 |
+
"step": 470
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.09,
|
| 294 |
+
"learning_rate": 0.0009951322345913224,
|
| 295 |
+
"loss": 2.1133,
|
| 296 |
+
"step": 480
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.09,
|
| 300 |
+
"learning_rate": 0.0009949276454812408,
|
| 301 |
+
"loss": 2.0844,
|
| 302 |
+
"step": 490
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.09,
|
| 306 |
+
"learning_rate": 0.00099471886694533,
|
| 307 |
+
"loss": 2.0796,
|
| 308 |
+
"step": 500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.09,
|
| 312 |
+
"learning_rate": 0.0009945059007508434,
|
| 313 |
+
"loss": 2.1255,
|
| 314 |
+
"step": 510
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.1,
|
| 318 |
+
"learning_rate": 0.0009942887487004804,
|
| 319 |
+
"loss": 2.0913,
|
| 320 |
+
"step": 520
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.1,
|
| 324 |
+
"learning_rate": 0.0009940674126323733,
|
| 325 |
+
"loss": 2.1003,
|
| 326 |
+
"step": 530
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.1,
|
| 330 |
+
"learning_rate": 0.0009938418944200709,
|
| 331 |
+
"loss": 2.0541,
|
| 332 |
+
"step": 540
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.1,
|
| 336 |
+
"learning_rate": 0.0009936121959725223,
|
| 337 |
+
"loss": 2.0523,
|
| 338 |
+
"step": 550
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.1,
|
| 342 |
+
"learning_rate": 0.0009933783192340618,
|
| 343 |
+
"loss": 2.1225,
|
| 344 |
+
"step": 560
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.11,
|
| 348 |
+
"learning_rate": 0.0009931402661843911,
|
| 349 |
+
"loss": 2.0446,
|
| 350 |
+
"step": 570
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.11,
|
| 354 |
+
"learning_rate": 0.000992898038838564,
|
| 355 |
+
"loss": 2.0921,
|
| 356 |
+
"step": 580
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.11,
|
| 360 |
+
"learning_rate": 0.0009926516392469674,
|
| 361 |
+
"loss": 2.1081,
|
| 362 |
+
"step": 590
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.11,
|
| 366 |
+
"learning_rate": 0.0009924010694953064,
|
| 367 |
+
"loss": 2.0734,
|
| 368 |
+
"step": 600
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.11,
|
| 372 |
+
"learning_rate": 0.0009921463317045843,
|
| 373 |
+
"loss": 2.0652,
|
| 374 |
+
"step": 610
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.11,
|
| 378 |
+
"learning_rate": 0.0009918874280310862,
|
| 379 |
+
"loss": 2.0818,
|
| 380 |
+
"step": 620
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.12,
|
| 384 |
+
"learning_rate": 0.0009916243606663605,
|
| 385 |
+
"loss": 2.0776,
|
| 386 |
+
"step": 630
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.12,
|
| 390 |
+
"learning_rate": 0.0009913571318371994,
|
| 391 |
+
"loss": 2.1025,
|
| 392 |
+
"step": 640
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.12,
|
| 396 |
+
"learning_rate": 0.0009910857438056215,
|
| 397 |
+
"loss": 2.066,
|
| 398 |
+
"step": 650
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.12,
|
| 402 |
+
"learning_rate": 0.0009908101988688512,
|
| 403 |
+
"loss": 2.0575,
|
| 404 |
+
"step": 660
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.12,
|
| 408 |
+
"learning_rate": 0.0009905304993593008,
|
| 409 |
+
"loss": 2.1269,
|
| 410 |
+
"step": 670
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.13,
|
| 414 |
+
"learning_rate": 0.0009902466476445486,
|
| 415 |
+
"loss": 2.0518,
|
| 416 |
+
"step": 680
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.13,
|
| 420 |
+
"learning_rate": 0.0009899586461273218,
|
| 421 |
+
"loss": 2.0698,
|
| 422 |
+
"step": 690
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.13,
|
| 426 |
+
"learning_rate": 0.000989666497245473,
|
| 427 |
+
"loss": 2.0988,
|
| 428 |
+
"step": 700
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.13,
|
| 432 |
+
"learning_rate": 0.0009893702034719624,
|
| 433 |
+
"loss": 2.0986,
|
| 434 |
+
"step": 710
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.13,
|
| 438 |
+
"learning_rate": 0.0009890697673148345,
|
| 439 |
+
"loss": 2.0237,
|
| 440 |
+
"step": 720
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.14,
|
| 444 |
+
"learning_rate": 0.0009887651913171986,
|
| 445 |
+
"loss": 2.0027,
|
| 446 |
+
"step": 730
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.14,
|
| 450 |
+
"learning_rate": 0.0009884564780572064,
|
| 451 |
+
"loss": 2.0563,
|
| 452 |
+
"step": 740
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.14,
|
| 456 |
+
"learning_rate": 0.0009881436301480305,
|
| 457 |
+
"loss": 2.0624,
|
| 458 |
+
"step": 750
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.14,
|
| 462 |
+
"learning_rate": 0.000987826650237842,
|
| 463 |
+
"loss": 2.0926,
|
| 464 |
+
"step": 760
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.14,
|
| 468 |
+
"learning_rate": 0.000987505541009788,
|
| 469 |
+
"loss": 2.0585,
|
| 470 |
+
"step": 770
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.14,
|
| 474 |
+
"learning_rate": 0.0009871803051819696,
|
| 475 |
+
"loss": 2.0494,
|
| 476 |
+
"step": 780
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.15,
|
| 480 |
+
"learning_rate": 0.0009868509455074183,
|
| 481 |
+
"loss": 2.0106,
|
| 482 |
+
"step": 790
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.15,
|
| 486 |
+
"learning_rate": 0.0009865174647740729,
|
| 487 |
+
"loss": 2.0861,
|
| 488 |
+
"step": 800
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.15,
|
| 492 |
+
"learning_rate": 0.0009861798658047556,
|
| 493 |
+
"loss": 2.0478,
|
| 494 |
+
"step": 810
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.15,
|
| 498 |
+
"learning_rate": 0.0009858381514571484,
|
| 499 |
+
"loss": 2.0469,
|
| 500 |
+
"step": 820
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.15,
|
| 504 |
+
"learning_rate": 0.000985492324623769,
|
| 505 |
+
"loss": 2.0671,
|
| 506 |
+
"step": 830
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.16,
|
| 510 |
+
"learning_rate": 0.0009851423882319458,
|
| 511 |
+
"loss": 2.0808,
|
| 512 |
+
"step": 840
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.16,
|
| 516 |
+
"learning_rate": 0.0009847883452437937,
|
| 517 |
+
"loss": 2.0331,
|
| 518 |
+
"step": 850
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.16,
|
| 522 |
+
"learning_rate": 0.0009844301986561893,
|
| 523 |
+
"loss": 2.0295,
|
| 524 |
+
"step": 860
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.16,
|
| 528 |
+
"learning_rate": 0.000984067951500744,
|
| 529 |
+
"loss": 2.0873,
|
| 530 |
+
"step": 870
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.16,
|
| 534 |
+
"learning_rate": 0.00098370160684378,
|
| 535 |
+
"loss": 2.1038,
|
| 536 |
+
"step": 880
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.16,
|
| 540 |
+
"learning_rate": 0.0009833311677863042,
|
| 541 |
+
"loss": 2.0337,
|
| 542 |
+
"step": 890
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.17,
|
| 546 |
+
"learning_rate": 0.0009829566374639801,
|
| 547 |
+
"loss": 2.0407,
|
| 548 |
+
"step": 900
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.17,
|
| 552 |
+
"learning_rate": 0.0009825780190471042,
|
| 553 |
+
"loss": 2.1049,
|
| 554 |
+
"step": 910
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.17,
|
| 558 |
+
"learning_rate": 0.000982195315740576,
|
| 559 |
+
"loss": 2.0475,
|
| 560 |
+
"step": 920
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.17,
|
| 564 |
+
"learning_rate": 0.0009818085307838741,
|
| 565 |
+
"loss": 2.0624,
|
| 566 |
+
"step": 930
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.17,
|
| 570 |
+
"learning_rate": 0.000981417667451026,
|
| 571 |
+
"loss": 2.0714,
|
| 572 |
+
"step": 940
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.18,
|
| 576 |
+
"learning_rate": 0.0009810227290505816,
|
| 577 |
+
"loss": 2.0947,
|
| 578 |
+
"step": 950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.18,
|
| 582 |
+
"learning_rate": 0.0009806237189255859,
|
| 583 |
+
"loss": 2.0591,
|
| 584 |
+
"step": 960
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.18,
|
| 588 |
+
"learning_rate": 0.0009802206404535489,
|
| 589 |
+
"loss": 2.0301,
|
| 590 |
+
"step": 970
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.18,
|
| 594 |
+
"learning_rate": 0.000979813497046419,
|
| 595 |
+
"loss": 2.0556,
|
| 596 |
+
"step": 980
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.18,
|
| 600 |
+
"learning_rate": 0.0009794022921505523,
|
| 601 |
+
"loss": 2.0753,
|
| 602 |
+
"step": 990
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.19,
|
| 606 |
+
"learning_rate": 0.000978987029246685,
|
| 607 |
+
"loss": 2.0898,
|
| 608 |
+
"step": 1000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.19,
|
| 612 |
+
"learning_rate": 0.0009785677118499029,
|
| 613 |
+
"loss": 2.0464,
|
| 614 |
+
"step": 1010
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.19,
|
| 618 |
+
"learning_rate": 0.0009781443435096116,
|
| 619 |
+
"loss": 2.0828,
|
| 620 |
+
"step": 1020
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.19,
|
| 624 |
+
"learning_rate": 0.0009777169278095074,
|
| 625 |
+
"loss": 2.1137,
|
| 626 |
+
"step": 1030
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.19,
|
| 630 |
+
"learning_rate": 0.0009772854683675462,
|
| 631 |
+
"loss": 2.0167,
|
| 632 |
+
"step": 1040
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.19,
|
| 636 |
+
"learning_rate": 0.000976849968835913,
|
| 637 |
+
"loss": 2.07,
|
| 638 |
+
"step": 1050
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 0.2,
|
| 642 |
+
"learning_rate": 0.0009764104329009909,
|
| 643 |
+
"loss": 2.0409,
|
| 644 |
+
"step": 1060
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.2,
|
| 648 |
+
"learning_rate": 0.0009759668642833304,
|
| 649 |
+
"loss": 2.015,
|
| 650 |
+
"step": 1070
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.2,
|
| 654 |
+
"learning_rate": 0.0009755192667376173,
|
| 655 |
+
"loss": 2.0175,
|
| 656 |
+
"step": 1080
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.2,
|
| 660 |
+
"learning_rate": 0.0009750676440526411,
|
| 661 |
+
"loss": 2.0773,
|
| 662 |
+
"step": 1090
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.2,
|
| 666 |
+
"learning_rate": 0.0009746120000512632,
|
| 667 |
+
"loss": 2.0245,
|
| 668 |
+
"step": 1100
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.21,
|
| 672 |
+
"learning_rate": 0.0009741523385903841,
|
| 673 |
+
"loss": 2.094,
|
| 674 |
+
"step": 1110
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.21,
|
| 678 |
+
"learning_rate": 0.0009736886635609112,
|
| 679 |
+
"loss": 2.0506,
|
| 680 |
+
"step": 1120
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.21,
|
| 684 |
+
"learning_rate": 0.0009732209788877258,
|
| 685 |
+
"loss": 2.0287,
|
| 686 |
+
"step": 1130
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 0.21,
|
| 690 |
+
"learning_rate": 0.0009727492885296489,
|
| 691 |
+
"loss": 2.1162,
|
| 692 |
+
"step": 1140
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.21,
|
| 696 |
+
"learning_rate": 0.0009722735964794099,
|
| 697 |
+
"loss": 2.1096,
|
| 698 |
+
"step": 1150
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.21,
|
| 702 |
+
"learning_rate": 0.0009717939067636099,
|
| 703 |
+
"loss": 2.0621,
|
| 704 |
+
"step": 1160
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.22,
|
| 708 |
+
"learning_rate": 0.0009713102234426903,
|
| 709 |
+
"loss": 2.0796,
|
| 710 |
+
"step": 1170
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.22,
|
| 714 |
+
"learning_rate": 0.0009708225506108965,
|
| 715 |
+
"loss": 2.0565,
|
| 716 |
+
"step": 1180
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.22,
|
| 720 |
+
"learning_rate": 0.0009703308923962447,
|
| 721 |
+
"loss": 2.0669,
|
| 722 |
+
"step": 1190
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.22,
|
| 726 |
+
"learning_rate": 0.0009698352529604857,
|
| 727 |
+
"loss": 2.0638,
|
| 728 |
+
"step": 1200
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 0.22,
|
| 732 |
+
"learning_rate": 0.0009693356364990705,
|
| 733 |
+
"loss": 2.0358,
|
| 734 |
+
"step": 1210
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 0.23,
|
| 738 |
+
"learning_rate": 0.0009688320472411143,
|
| 739 |
+
"loss": 2.0859,
|
| 740 |
+
"step": 1220
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.23,
|
| 744 |
+
"learning_rate": 0.0009683244894493613,
|
| 745 |
+
"loss": 2.0932,
|
| 746 |
+
"step": 1230
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.23,
|
| 750 |
+
"learning_rate": 0.0009678129674201479,
|
| 751 |
+
"loss": 2.0129,
|
| 752 |
+
"step": 1240
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.23,
|
| 756 |
+
"learning_rate": 0.0009672974854833669,
|
| 757 |
+
"loss": 2.055,
|
| 758 |
+
"step": 1250
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.23,
|
| 762 |
+
"learning_rate": 0.0009667780480024304,
|
| 763 |
+
"loss": 2.0665,
|
| 764 |
+
"step": 1260
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.24,
|
| 768 |
+
"learning_rate": 0.0009662546593742334,
|
| 769 |
+
"loss": 2.0488,
|
| 770 |
+
"step": 1270
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 0.24,
|
| 774 |
+
"learning_rate": 0.0009657273240291159,
|
| 775 |
+
"loss": 2.0543,
|
| 776 |
+
"step": 1280
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.24,
|
| 780 |
+
"learning_rate": 0.0009651960464308261,
|
| 781 |
+
"loss": 2.0418,
|
| 782 |
+
"step": 1290
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.24,
|
| 786 |
+
"learning_rate": 0.0009646608310764819,
|
| 787 |
+
"loss": 2.033,
|
| 788 |
+
"step": 1300
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.24,
|
| 792 |
+
"learning_rate": 0.0009641216824965338,
|
| 793 |
+
"loss": 2.1034,
|
| 794 |
+
"step": 1310
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.24,
|
| 798 |
+
"learning_rate": 0.0009635786052547253,
|
| 799 |
+
"loss": 2.0866,
|
| 800 |
+
"step": 1320
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.25,
|
| 804 |
+
"learning_rate": 0.0009630316039480556,
|
| 805 |
+
"loss": 2.0607,
|
| 806 |
+
"step": 1330
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.25,
|
| 810 |
+
"learning_rate": 0.0009624806832067394,
|
| 811 |
+
"loss": 2.0457,
|
| 812 |
+
"step": 1340
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.25,
|
| 816 |
+
"learning_rate": 0.0009619258476941686,
|
| 817 |
+
"loss": 2.032,
|
| 818 |
+
"step": 1350
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.25,
|
| 822 |
+
"learning_rate": 0.000961367102106873,
|
| 823 |
+
"loss": 2.0519,
|
| 824 |
+
"step": 1360
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.25,
|
| 828 |
+
"learning_rate": 0.0009608044511744791,
|
| 829 |
+
"loss": 2.0449,
|
| 830 |
+
"step": 1370
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.26,
|
| 834 |
+
"learning_rate": 0.0009602378996596721,
|
| 835 |
+
"loss": 1.9949,
|
| 836 |
+
"step": 1380
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.26,
|
| 840 |
+
"learning_rate": 0.0009596674523581539,
|
| 841 |
+
"loss": 2.0394,
|
| 842 |
+
"step": 1390
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.26,
|
| 846 |
+
"learning_rate": 0.0009590931140986035,
|
| 847 |
+
"loss": 2.0386,
|
| 848 |
+
"step": 1400
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.26,
|
| 852 |
+
"learning_rate": 0.0009585148897426354,
|
| 853 |
+
"loss": 2.0254,
|
| 854 |
+
"step": 1410
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.26,
|
| 858 |
+
"learning_rate": 0.0009579327841847593,
|
| 859 |
+
"loss": 2.0238,
|
| 860 |
+
"step": 1420
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.26,
|
| 864 |
+
"learning_rate": 0.000957346802352338,
|
| 865 |
+
"loss": 2.0509,
|
| 866 |
+
"step": 1430
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.27,
|
| 870 |
+
"learning_rate": 0.0009567569492055456,
|
| 871 |
+
"loss": 2.0004,
|
| 872 |
+
"step": 1440
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.27,
|
| 876 |
+
"learning_rate": 0.0009561632297373263,
|
| 877 |
+
"loss": 2.0203,
|
| 878 |
+
"step": 1450
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.27,
|
| 882 |
+
"learning_rate": 0.0009555656489733513,
|
| 883 |
+
"loss": 2.0182,
|
| 884 |
+
"step": 1460
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.27,
|
| 888 |
+
"learning_rate": 0.000954964211971977,
|
| 889 |
+
"loss": 1.9754,
|
| 890 |
+
"step": 1470
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.27,
|
| 894 |
+
"learning_rate": 0.0009543589238242012,
|
| 895 |
+
"loss": 2.0374,
|
| 896 |
+
"step": 1480
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.28,
|
| 900 |
+
"learning_rate": 0.000953749789653621,
|
| 901 |
+
"loss": 2.0367,
|
| 902 |
+
"step": 1490
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.28,
|
| 906 |
+
"learning_rate": 0.000953136814616389,
|
| 907 |
+
"loss": 2.0866,
|
| 908 |
+
"step": 1500
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.28,
|
| 912 |
+
"learning_rate": 0.0009525200039011694,
|
| 913 |
+
"loss": 2.0083,
|
| 914 |
+
"step": 1510
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.28,
|
| 918 |
+
"learning_rate": 0.0009518993627290948,
|
| 919 |
+
"loss": 2.0525,
|
| 920 |
+
"step": 1520
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.28,
|
| 924 |
+
"learning_rate": 0.0009512748963537212,
|
| 925 |
+
"loss": 2.0636,
|
| 926 |
+
"step": 1530
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.29,
|
| 930 |
+
"learning_rate": 0.000950646610060984,
|
| 931 |
+
"loss": 2.0522,
|
| 932 |
+
"step": 1540
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.29,
|
| 936 |
+
"learning_rate": 0.0009500145091691532,
|
| 937 |
+
"loss": 2.05,
|
| 938 |
+
"step": 1550
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 0.29,
|
| 942 |
+
"learning_rate": 0.0009493785990287882,
|
| 943 |
+
"loss": 1.9887,
|
| 944 |
+
"step": 1560
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 0.29,
|
| 948 |
+
"learning_rate": 0.0009487388850226926,
|
| 949 |
+
"loss": 2.0309,
|
| 950 |
+
"step": 1570
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.29,
|
| 954 |
+
"learning_rate": 0.000948095372565869,
|
| 955 |
+
"loss": 1.9954,
|
| 956 |
+
"step": 1580
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 0.29,
|
| 960 |
+
"learning_rate": 0.0009474480671054726,
|
| 961 |
+
"loss": 2.0078,
|
| 962 |
+
"step": 1590
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.3,
|
| 966 |
+
"learning_rate": 0.0009467969741207652,
|
| 967 |
+
"loss": 2.0395,
|
| 968 |
+
"step": 1600
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.3,
|
| 972 |
+
"learning_rate": 0.0009461420991230693,
|
| 973 |
+
"loss": 2.0415,
|
| 974 |
+
"step": 1610
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 0.3,
|
| 978 |
+
"learning_rate": 0.0009454834476557207,
|
| 979 |
+
"loss": 2.0308,
|
| 980 |
+
"step": 1620
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 0.3,
|
| 984 |
+
"learning_rate": 0.0009448210252940223,
|
| 985 |
+
"loss": 2.0826,
|
| 986 |
+
"step": 1630
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.3,
|
| 990 |
+
"learning_rate": 0.0009441548376451963,
|
| 991 |
+
"loss": 2.0424,
|
| 992 |
+
"step": 1640
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.31,
|
| 996 |
+
"learning_rate": 0.0009434848903483373,
|
| 997 |
+
"loss": 2.0125,
|
| 998 |
+
"step": 1650
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 0.31,
|
| 1002 |
+
"learning_rate": 0.0009428111890743639,
|
| 1003 |
+
"loss": 2.0139,
|
| 1004 |
+
"step": 1660
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.31,
|
| 1008 |
+
"learning_rate": 0.0009421337395259717,
|
| 1009 |
+
"loss": 2.0682,
|
| 1010 |
+
"step": 1670
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.31,
|
| 1014 |
+
"learning_rate": 0.0009414525474375837,
|
| 1015 |
+
"loss": 2.0577,
|
| 1016 |
+
"step": 1680
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 0.31,
|
| 1020 |
+
"learning_rate": 0.0009407676185753029,
|
| 1021 |
+
"loss": 2.0262,
|
| 1022 |
+
"step": 1690
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 0.31,
|
| 1026 |
+
"learning_rate": 0.0009400789587368632,
|
| 1027 |
+
"loss": 2.0515,
|
| 1028 |
+
"step": 1700
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 0.32,
|
| 1032 |
+
"learning_rate": 0.0009393865737515794,
|
| 1033 |
+
"loss": 2.0398,
|
| 1034 |
+
"step": 1710
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 0.32,
|
| 1038 |
+
"learning_rate": 0.0009386904694802997,
|
| 1039 |
+
"loss": 2.0146,
|
| 1040 |
+
"step": 1720
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.32,
|
| 1044 |
+
"learning_rate": 0.0009379906518153543,
|
| 1045 |
+
"loss": 2.0438,
|
| 1046 |
+
"step": 1730
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.32,
|
| 1050 |
+
"learning_rate": 0.0009372871266805063,
|
| 1051 |
+
"loss": 2.0377,
|
| 1052 |
+
"step": 1740
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.32,
|
| 1056 |
+
"learning_rate": 0.000936579900030902,
|
| 1057 |
+
"loss": 2.0789,
|
| 1058 |
+
"step": 1750
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 0.33,
|
| 1062 |
+
"learning_rate": 0.0009358689778530193,
|
| 1063 |
+
"loss": 2.0201,
|
| 1064 |
+
"step": 1760
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 0.33,
|
| 1068 |
+
"learning_rate": 0.0009351543661646185,
|
| 1069 |
+
"loss": 2.0114,
|
| 1070 |
+
"step": 1770
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.33,
|
| 1074 |
+
"learning_rate": 0.0009344360710146898,
|
| 1075 |
+
"loss": 2.0242,
|
| 1076 |
+
"step": 1780
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.33,
|
| 1080 |
+
"learning_rate": 0.0009337140984834034,
|
| 1081 |
+
"loss": 2.0436,
|
| 1082 |
+
"step": 1790
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.33,
|
| 1086 |
+
"learning_rate": 0.0009329884546820572,
|
| 1087 |
+
"loss": 2.0452,
|
| 1088 |
+
"step": 1800
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.34,
|
| 1092 |
+
"learning_rate": 0.000932259145753026,
|
| 1093 |
+
"loss": 2.0254,
|
| 1094 |
+
"step": 1810
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.34,
|
| 1098 |
+
"learning_rate": 0.0009315261778697083,
|
| 1099 |
+
"loss": 2.0409,
|
| 1100 |
+
"step": 1820
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 0.34,
|
| 1104 |
+
"learning_rate": 0.0009307895572364746,
|
| 1105 |
+
"loss": 2.0301,
|
| 1106 |
+
"step": 1830
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 0.34,
|
| 1110 |
+
"learning_rate": 0.0009300492900886154,
|
| 1111 |
+
"loss": 2.0078,
|
| 1112 |
+
"step": 1840
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.34,
|
| 1116 |
+
"learning_rate": 0.0009293053826922873,
|
| 1117 |
+
"loss": 1.9851,
|
| 1118 |
+
"step": 1850
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.34,
|
| 1122 |
+
"learning_rate": 0.0009285578413444613,
|
| 1123 |
+
"loss": 1.9947,
|
| 1124 |
+
"step": 1860
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.35,
|
| 1128 |
+
"learning_rate": 0.0009278066723728682,
|
| 1129 |
+
"loss": 2.0331,
|
| 1130 |
+
"step": 1870
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.35,
|
| 1134 |
+
"learning_rate": 0.0009270518821359461,
|
| 1135 |
+
"loss": 2.0058,
|
| 1136 |
+
"step": 1880
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.35,
|
| 1140 |
+
"learning_rate": 0.0009262934770227858,
|
| 1141 |
+
"loss": 2.05,
|
| 1142 |
+
"step": 1890
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 0.35,
|
| 1146 |
+
"learning_rate": 0.0009255314634530771,
|
| 1147 |
+
"loss": 2.0444,
|
| 1148 |
+
"step": 1900
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 0.35,
|
| 1152 |
+
"learning_rate": 0.0009247658478770543,
|
| 1153 |
+
"loss": 2.0045,
|
| 1154 |
+
"step": 1910
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.36,
|
| 1158 |
+
"learning_rate": 0.000923996636775442,
|
| 1159 |
+
"loss": 2.0211,
|
| 1160 |
+
"step": 1920
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.36,
|
| 1164 |
+
"learning_rate": 0.0009232238366593997,
|
| 1165 |
+
"loss": 2.0124,
|
| 1166 |
+
"step": 1930
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 0.36,
|
| 1170 |
+
"learning_rate": 0.0009224474540704671,
|
| 1171 |
+
"loss": 2.0067,
|
| 1172 |
+
"step": 1940
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.36,
|
| 1176 |
+
"learning_rate": 0.0009216674955805079,
|
| 1177 |
+
"loss": 2.0247,
|
| 1178 |
+
"step": 1950
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.36,
|
| 1182 |
+
"learning_rate": 0.0009208839677916557,
|
| 1183 |
+
"loss": 2.0314,
|
| 1184 |
+
"step": 1960
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 0.36,
|
| 1188 |
+
"learning_rate": 0.0009200968773362568,
|
| 1189 |
+
"loss": 2.067,
|
| 1190 |
+
"step": 1970
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 0.37,
|
| 1194 |
+
"learning_rate": 0.0009193062308768145,
|
| 1195 |
+
"loss": 2.0168,
|
| 1196 |
+
"step": 1980
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.37,
|
| 1200 |
+
"learning_rate": 0.0009185120351059326,
|
| 1201 |
+
"loss": 2.0649,
|
| 1202 |
+
"step": 1990
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.37,
|
| 1206 |
+
"learning_rate": 0.0009177142967462591,
|
| 1207 |
+
"loss": 2.0208,
|
| 1208 |
+
"step": 2000
|
| 1209 |
+
}
|
| 1210 |
+
],
|
| 1211 |
+
"max_steps": 10798,
|
| 1212 |
+
"num_train_epochs": 2,
|
| 1213 |
+
"total_flos": 1.8983090426321306e+18,
|
| 1214 |
+
"trial_name": null,
|
| 1215 |
+
"trial_params": null
|
| 1216 |
+
}
|
checkpoint-2000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8813832b030b0a678cb3ad0abd28db5000bdb8bb1708e45ba10cfda908fcc38
|
| 3 |
+
size 3305
|
checkpoint-3000/README.md
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
checkpoint-3000/adapter_config.json
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"base_model_name_or_path": "chatglm2-6b",
|
| 3 |
+
"bias": "none",
|
| 4 |
+
"fan_in_fan_out": false,
|
| 5 |
+
"inference_mode": true,
|
| 6 |
+
"init_lora_weights": true,
|
| 7 |
+
"layers_pattern": null,
|
| 8 |
+
"layers_to_transform": null,
|
| 9 |
+
"lora_alpha": 32.0,
|
| 10 |
+
"lora_dropout": 0.1,
|
| 11 |
+
"modules_to_save": null,
|
| 12 |
+
"peft_type": "LORA",
|
| 13 |
+
"r": 8,
|
| 14 |
+
"revision": null,
|
| 15 |
+
"target_modules": [
|
| 16 |
+
"query_key_value"
|
| 17 |
+
],
|
| 18 |
+
"task_type": "CAUSAL_LM"
|
| 19 |
+
}
|
checkpoint-3000/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a3cd0c710c3e6e522341f9bc18338a3a5d364603e39d5a7138ce0ece693cada4
|
| 3 |
+
size 7819417
|
checkpoint-3000/finetuning_args.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"finetuning_type": "lora",
|
| 3 |
+
"lora_alpha": 32.0,
|
| 4 |
+
"lora_dropout": 0.1,
|
| 5 |
+
"lora_rank": 8,
|
| 6 |
+
"lora_target": [
|
| 7 |
+
"query_key_value"
|
| 8 |
+
],
|
| 9 |
+
"name_module_trainable": "mlp",
|
| 10 |
+
"num_layer_trainable": 3,
|
| 11 |
+
"pre_seq_len": 64,
|
| 12 |
+
"prefix_projection": false
|
| 13 |
+
}
|
checkpoint-3000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2286a7eb891d1c724b052fc718f53e053668dd120909c7d17063e0b11f1046c7
|
| 3 |
+
size 15644485
|
checkpoint-3000/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:28df0ebee892c2cd596ed2d9c3c1b1bce8aacce528339850151675cb3e979e11
|
| 3 |
+
size 18679
|
checkpoint-3000/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ff24f859606279e5f5564c5772b357bdf34d74d1fb9f644a200359edb9d5a2bb
|
| 3 |
+
size 18679
|
checkpoint-3000/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4cf19f9e07dfb1c6a4f9b043721d057abeb2a702aa4992a0d2dab15224d64805
|
| 3 |
+
size 18679
|
checkpoint-3000/rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:723718e2b5b73a3d4bfee55008ffbae341c90ae5f2480575806ffc1134c69a8e
|
| 3 |
+
size 18679
|
checkpoint-3000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9cf9b363bf12cc68bc5c1a58285c6f89fc38fb83c9837ee26212120079972ec5
|
| 3 |
+
size 627
|