Upload 8 files
Browse files- adapter_config.json +37 -0
- adapter_model.safetensors +3 -0
- additional_config.json +1 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +2181 -0
- training_args.bin +3 -0
adapter_config.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "Qwen2.5-32B-Instruct",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"eva_config": null,
|
| 7 |
+
"exclude_modules": null,
|
| 8 |
+
"fan_in_fan_out": false,
|
| 9 |
+
"inference_mode": true,
|
| 10 |
+
"init_lora_weights": true,
|
| 11 |
+
"layer_replication": null,
|
| 12 |
+
"layers_pattern": null,
|
| 13 |
+
"layers_to_transform": null,
|
| 14 |
+
"loftq_config": {},
|
| 15 |
+
"lora_alpha": 32,
|
| 16 |
+
"lora_bias": false,
|
| 17 |
+
"lora_dropout": 0.0,
|
| 18 |
+
"megatron_config": null,
|
| 19 |
+
"megatron_core": "megatron.core",
|
| 20 |
+
"modules_to_save": null,
|
| 21 |
+
"peft_type": "LORA",
|
| 22 |
+
"r": 16,
|
| 23 |
+
"rank_pattern": {},
|
| 24 |
+
"revision": null,
|
| 25 |
+
"target_modules": [
|
| 26 |
+
"q_proj",
|
| 27 |
+
"up_proj",
|
| 28 |
+
"k_proj",
|
| 29 |
+
"o_proj",
|
| 30 |
+
"gate_proj",
|
| 31 |
+
"down_proj",
|
| 32 |
+
"v_proj"
|
| 33 |
+
],
|
| 34 |
+
"task_type": "CAUSAL_LM",
|
| 35 |
+
"use_dora": false,
|
| 36 |
+
"use_rslora": false
|
| 37 |
+
}
|
adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f07bba4c7af6aee9fc3ae96d9ad58ab8a96f35477a25c20e5acb115200364f2e
|
| 3 |
+
size 536991984
|
additional_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"lora_dtype": null, "lorap_lr_ratio": null, "lorap_emb_lr": 1e-06}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4b01ce863f589f77a7037018bbf8cd42f4aaf0d9b106655f95db8129103f89e7
|
| 3 |
+
size 1074499986
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e3e5d946241df2516b06d7074d8779088eae7607173ad780df56583910a9589b
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5414b7afe4dad39d837a6989445b96e11d5b91381e5d3d286fcfc6c89949500f
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2181 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.58375472,
|
| 3 |
+
"best_model_checkpoint": "/export/home2/zli/kc/mm_rag/Qwen2.5-32B-Instruct_lora/checkpoint-1026",
|
| 4 |
+
"epoch": 0.9997563946406821,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 1026,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.00097442143727162,
|
| 13 |
+
"grad_norm": 0.8529725074768066,
|
| 14 |
+
"learning_rate": 1.9230769230769234e-06,
|
| 15 |
+
"loss": 0.9744532108306885,
|
| 16 |
+
"memory(GiB)": 255.29,
|
| 17 |
+
"step": 1,
|
| 18 |
+
"token_acc": 0.7235609103078983,
|
| 19 |
+
"train_speed(iter/s)": 0.048136
|
| 20 |
+
},
|
| 21 |
+
{
|
| 22 |
+
"epoch": 0.0048721071863581,
|
| 23 |
+
"grad_norm": 0.9313930869102478,
|
| 24 |
+
"learning_rate": 9.615384615384616e-06,
|
| 25 |
+
"loss": 0.7735831141471863,
|
| 26 |
+
"memory(GiB)": 307.13,
|
| 27 |
+
"step": 5,
|
| 28 |
+
"token_acc": 0.7883002659030477,
|
| 29 |
+
"train_speed(iter/s)": 0.052895
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 0.0097442143727162,
|
| 33 |
+
"grad_norm": 0.5147601962089539,
|
| 34 |
+
"learning_rate": 1.923076923076923e-05,
|
| 35 |
+
"loss": 0.8932640075683593,
|
| 36 |
+
"memory(GiB)": 357.61,
|
| 37 |
+
"step": 10,
|
| 38 |
+
"token_acc": 0.7670103092783506,
|
| 39 |
+
"train_speed(iter/s)": 0.055392
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.014616321559074299,
|
| 43 |
+
"grad_norm": 0.3254891037940979,
|
| 44 |
+
"learning_rate": 2.8846153846153845e-05,
|
| 45 |
+
"loss": 0.7951784610748291,
|
| 46 |
+
"memory(GiB)": 357.63,
|
| 47 |
+
"step": 15,
|
| 48 |
+
"token_acc": 0.7827709279688514,
|
| 49 |
+
"train_speed(iter/s)": 0.058176
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.0194884287454324,
|
| 53 |
+
"grad_norm": 0.5581763982772827,
|
| 54 |
+
"learning_rate": 3.846153846153846e-05,
|
| 55 |
+
"loss": 0.956171703338623,
|
| 56 |
+
"memory(GiB)": 357.63,
|
| 57 |
+
"step": 20,
|
| 58 |
+
"token_acc": 0.7419072615923009,
|
| 59 |
+
"train_speed(iter/s)": 0.05832
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.024360535931790498,
|
| 63 |
+
"grad_norm": 0.3584084212779999,
|
| 64 |
+
"learning_rate": 4.8076923076923084e-05,
|
| 65 |
+
"loss": 0.8373254776000977,
|
| 66 |
+
"memory(GiB)": 357.64,
|
| 67 |
+
"step": 25,
|
| 68 |
+
"token_acc": 0.7718067293892521,
|
| 69 |
+
"train_speed(iter/s)": 0.057013
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.029232643118148598,
|
| 73 |
+
"grad_norm": 0.43336302042007446,
|
| 74 |
+
"learning_rate": 5.769230769230769e-05,
|
| 75 |
+
"loss": 0.7370581150054931,
|
| 76 |
+
"memory(GiB)": 357.64,
|
| 77 |
+
"step": 30,
|
| 78 |
+
"token_acc": 0.7800821355236139,
|
| 79 |
+
"train_speed(iter/s)": 0.057318
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.0341047503045067,
|
| 83 |
+
"grad_norm": 0.6522459983825684,
|
| 84 |
+
"learning_rate": 6.730769230769232e-05,
|
| 85 |
+
"loss": 0.8316493988037109,
|
| 86 |
+
"memory(GiB)": 357.64,
|
| 87 |
+
"step": 35,
|
| 88 |
+
"token_acc": 0.7533095188064719,
|
| 89 |
+
"train_speed(iter/s)": 0.0583
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.0389768574908648,
|
| 93 |
+
"grad_norm": 0.2728117108345032,
|
| 94 |
+
"learning_rate": 7.692307692307693e-05,
|
| 95 |
+
"loss": 0.7739034652709961,
|
| 96 |
+
"memory(GiB)": 357.64,
|
| 97 |
+
"step": 40,
|
| 98 |
+
"token_acc": 0.7733619763694952,
|
| 99 |
+
"train_speed(iter/s)": 0.058058
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.0438489646772229,
|
| 103 |
+
"grad_norm": 0.5565893054008484,
|
| 104 |
+
"learning_rate": 8.653846153846155e-05,
|
| 105 |
+
"loss": 0.664579200744629,
|
| 106 |
+
"memory(GiB)": 357.64,
|
| 107 |
+
"step": 45,
|
| 108 |
+
"token_acc": 0.7940859054608245,
|
| 109 |
+
"train_speed(iter/s)": 0.058044
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.048721071863580996,
|
| 113 |
+
"grad_norm": 0.8710922002792358,
|
| 114 |
+
"learning_rate": 9.615384615384617e-05,
|
| 115 |
+
"loss": 0.6327468872070312,
|
| 116 |
+
"memory(GiB)": 378.13,
|
| 117 |
+
"step": 50,
|
| 118 |
+
"token_acc": 0.8048907388137357,
|
| 119 |
+
"train_speed(iter/s)": 0.057666
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.0535931790499391,
|
| 123 |
+
"grad_norm": 0.5048324465751648,
|
| 124 |
+
"learning_rate": 9.999765921804365e-05,
|
| 125 |
+
"loss": 0.7563381671905518,
|
| 126 |
+
"memory(GiB)": 378.13,
|
| 127 |
+
"step": 55,
|
| 128 |
+
"token_acc": 0.7769541778975741,
|
| 129 |
+
"train_speed(iter/s)": 0.056465
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.058465286236297195,
|
| 133 |
+
"grad_norm": 0.4348597824573517,
|
| 134 |
+
"learning_rate": 9.998335523311734e-05,
|
| 135 |
+
"loss": 0.6969810009002686,
|
| 136 |
+
"memory(GiB)": 378.13,
|
| 137 |
+
"step": 60,
|
| 138 |
+
"token_acc": 0.7881806108897742,
|
| 139 |
+
"train_speed(iter/s)": 0.056264
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 0.06333739342265529,
|
| 143 |
+
"grad_norm": 0.3995300233364105,
|
| 144 |
+
"learning_rate": 9.995605141340247e-05,
|
| 145 |
+
"loss": 0.6338334083557129,
|
| 146 |
+
"memory(GiB)": 378.13,
|
| 147 |
+
"step": 65,
|
| 148 |
+
"token_acc": 0.8006442376521117,
|
| 149 |
+
"train_speed(iter/s)": 0.056356
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.0682095006090134,
|
| 153 |
+
"grad_norm": 0.25958451628685,
|
| 154 |
+
"learning_rate": 9.991575486016592e-05,
|
| 155 |
+
"loss": 0.7143070697784424,
|
| 156 |
+
"memory(GiB)": 378.13,
|
| 157 |
+
"step": 70,
|
| 158 |
+
"token_acc": 0.7797544920832592,
|
| 159 |
+
"train_speed(iter/s)": 0.056342
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.0730816077953715,
|
| 163 |
+
"grad_norm": 0.3476051688194275,
|
| 164 |
+
"learning_rate": 9.986247605386727e-05,
|
| 165 |
+
"loss": 0.6742859840393066,
|
| 166 |
+
"memory(GiB)": 378.13,
|
| 167 |
+
"step": 75,
|
| 168 |
+
"token_acc": 0.7907263751763046,
|
| 169 |
+
"train_speed(iter/s)": 0.056661
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.0779537149817296,
|
| 173 |
+
"grad_norm": 0.3681824505329132,
|
| 174 |
+
"learning_rate": 9.979622885143301e-05,
|
| 175 |
+
"loss": 0.5565629959106445,
|
| 176 |
+
"memory(GiB)": 378.13,
|
| 177 |
+
"step": 80,
|
| 178 |
+
"token_acc": 0.8259753593429158,
|
| 179 |
+
"train_speed(iter/s)": 0.056802
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.0828258221680877,
|
| 183 |
+
"grad_norm": 0.4626450836658478,
|
| 184 |
+
"learning_rate": 9.97170304826526e-05,
|
| 185 |
+
"loss": 0.6965714931488037,
|
| 186 |
+
"memory(GiB)": 378.13,
|
| 187 |
+
"step": 85,
|
| 188 |
+
"token_acc": 0.7883802169516455,
|
| 189 |
+
"train_speed(iter/s)": 0.057068
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.0876979293544458,
|
| 193 |
+
"grad_norm": 0.6139679551124573,
|
| 194 |
+
"learning_rate": 9.962490154569727e-05,
|
| 195 |
+
"loss": 0.670227336883545,
|
| 196 |
+
"memory(GiB)": 378.13,
|
| 197 |
+
"step": 90,
|
| 198 |
+
"token_acc": 0.7868181818181819,
|
| 199 |
+
"train_speed(iter/s)": 0.056985
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.0925700365408039,
|
| 203 |
+
"grad_norm": 0.3063270151615143,
|
| 204 |
+
"learning_rate": 9.95198660017628e-05,
|
| 205 |
+
"loss": 0.6900132179260254,
|
| 206 |
+
"memory(GiB)": 378.13,
|
| 207 |
+
"step": 95,
|
| 208 |
+
"token_acc": 0.7969001610305958,
|
| 209 |
+
"train_speed(iter/s)": 0.056931
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.09744214372716199,
|
| 213 |
+
"grad_norm": 0.6139111518859863,
|
| 214 |
+
"learning_rate": 9.940195116883755e-05,
|
| 215 |
+
"loss": 0.6138424396514892,
|
| 216 |
+
"memory(GiB)": 378.13,
|
| 217 |
+
"step": 100,
|
| 218 |
+
"token_acc": 0.8159609120521173,
|
| 219 |
+
"train_speed(iter/s)": 0.056746
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.09744214372716199,
|
| 223 |
+
"eval_loss": 0.7148731350898743,
|
| 224 |
+
"eval_runtime": 6.1133,
|
| 225 |
+
"eval_samples_per_second": 0.654,
|
| 226 |
+
"eval_steps_per_second": 0.654,
|
| 227 |
+
"step": 100
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.1023142509135201,
|
| 231 |
+
"grad_norm": 0.2960892617702484,
|
| 232 |
+
"learning_rate": 9.927118771459763e-05,
|
| 233 |
+
"loss": 0.7610847473144531,
|
| 234 |
+
"memory(GiB)": 378.13,
|
| 235 |
+
"step": 105,
|
| 236 |
+
"token_acc": 0.767438173747622,
|
| 237 |
+
"train_speed(iter/s)": 0.055274
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.1071863580998782,
|
| 241 |
+
"grad_norm": 0.35161978006362915,
|
| 242 |
+
"learning_rate": 9.91276096484306e-05,
|
| 243 |
+
"loss": 0.553467845916748,
|
| 244 |
+
"memory(GiB)": 378.13,
|
| 245 |
+
"step": 110,
|
| 246 |
+
"token_acc": 0.8268870335661953,
|
| 247 |
+
"train_speed(iter/s)": 0.055276
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.1120584652862363,
|
| 251 |
+
"grad_norm": 0.412120521068573,
|
| 252 |
+
"learning_rate": 9.897125431259033e-05,
|
| 253 |
+
"loss": 0.634190034866333,
|
| 254 |
+
"memory(GiB)": 378.13,
|
| 255 |
+
"step": 115,
|
| 256 |
+
"token_acc": 0.8068276436303081,
|
| 257 |
+
"train_speed(iter/s)": 0.055265
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.11693057247259439,
|
| 261 |
+
"grad_norm": 0.458766907453537,
|
| 262 |
+
"learning_rate": 9.880216237248481e-05,
|
| 263 |
+
"loss": 0.5070098400115967,
|
| 264 |
+
"memory(GiB)": 383.55,
|
| 265 |
+
"step": 120,
|
| 266 |
+
"token_acc": 0.8488372093023255,
|
| 267 |
+
"train_speed(iter/s)": 0.055088
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.1218026796589525,
|
| 271 |
+
"grad_norm": 0.26979780197143555,
|
| 272 |
+
"learning_rate": 9.862037780609986e-05,
|
| 273 |
+
"loss": 0.6117064476013183,
|
| 274 |
+
"memory(GiB)": 383.55,
|
| 275 |
+
"step": 125,
|
| 276 |
+
"token_acc": 0.8204016064257028,
|
| 277 |
+
"train_speed(iter/s)": 0.055201
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 0.12667478684531058,
|
| 281 |
+
"grad_norm": 0.33145567774772644,
|
| 282 |
+
"learning_rate": 9.842594789256103e-05,
|
| 283 |
+
"loss": 0.6480350017547607,
|
| 284 |
+
"memory(GiB)": 383.55,
|
| 285 |
+
"step": 130,
|
| 286 |
+
"token_acc": 0.7939520333680917,
|
| 287 |
+
"train_speed(iter/s)": 0.055341
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 0.1315468940316687,
|
| 291 |
+
"grad_norm": 0.29335150122642517,
|
| 292 |
+
"learning_rate": 9.821892319983726e-05,
|
| 293 |
+
"loss": 0.6102027893066406,
|
| 294 |
+
"memory(GiB)": 383.55,
|
| 295 |
+
"step": 135,
|
| 296 |
+
"token_acc": 0.8075084115459537,
|
| 297 |
+
"train_speed(iter/s)": 0.055381
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.1364190012180268,
|
| 301 |
+
"grad_norm": 0.34363335371017456,
|
| 302 |
+
"learning_rate": 9.799935757158891e-05,
|
| 303 |
+
"loss": 0.6652801513671875,
|
| 304 |
+
"memory(GiB)": 383.55,
|
| 305 |
+
"step": 140,
|
| 306 |
+
"token_acc": 0.7967960995995125,
|
| 307 |
+
"train_speed(iter/s)": 0.055397
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 0.14129110840438489,
|
| 311 |
+
"grad_norm": 1.3419567346572876,
|
| 312 |
+
"learning_rate": 9.776730811316394e-05,
|
| 313 |
+
"loss": 0.6267284393310547,
|
| 314 |
+
"memory(GiB)": 383.55,
|
| 315 |
+
"step": 145,
|
| 316 |
+
"token_acc": 0.8041536400178652,
|
| 317 |
+
"train_speed(iter/s)": 0.055716
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.146163215590743,
|
| 321 |
+
"grad_norm": 0.32478609681129456,
|
| 322 |
+
"learning_rate": 9.752283517674575e-05,
|
| 323 |
+
"loss": 0.5990486145019531,
|
| 324 |
+
"memory(GiB)": 383.55,
|
| 325 |
+
"step": 150,
|
| 326 |
+
"token_acc": 0.80878414568827,
|
| 327 |
+
"train_speed(iter/s)": 0.055975
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.1510353227771011,
|
| 331 |
+
"grad_norm": 0.33393779397010803,
|
| 332 |
+
"learning_rate": 9.72660023456566e-05,
|
| 333 |
+
"loss": 0.6312044143676758,
|
| 334 |
+
"memory(GiB)": 383.55,
|
| 335 |
+
"step": 155,
|
| 336 |
+
"token_acc": 0.8099249502220861,
|
| 337 |
+
"train_speed(iter/s)": 0.055469
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 0.1559074299634592,
|
| 341 |
+
"grad_norm": 0.39661872386932373,
|
| 342 |
+
"learning_rate": 9.699687641782067e-05,
|
| 343 |
+
"loss": 0.727474308013916,
|
| 344 |
+
"memory(GiB)": 383.55,
|
| 345 |
+
"step": 160,
|
| 346 |
+
"token_acc": 0.7668711656441718,
|
| 347 |
+
"train_speed(iter/s)": 0.055534
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 0.1607795371498173,
|
| 351 |
+
"grad_norm": 0.3174588978290558,
|
| 352 |
+
"learning_rate": 9.671552738839099e-05,
|
| 353 |
+
"loss": 0.7284453868865967,
|
| 354 |
+
"memory(GiB)": 383.55,
|
| 355 |
+
"step": 165,
|
| 356 |
+
"token_acc": 0.7788385043754972,
|
| 357 |
+
"train_speed(iter/s)": 0.055621
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.1656516443361754,
|
| 361 |
+
"grad_norm": 0.3508811891078949,
|
| 362 |
+
"learning_rate": 9.642202843154491e-05,
|
| 363 |
+
"loss": 0.6260187149047851,
|
| 364 |
+
"memory(GiB)": 383.55,
|
| 365 |
+
"step": 170,
|
| 366 |
+
"token_acc": 0.8036006546644845,
|
| 367 |
+
"train_speed(iter/s)": 0.055841
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.1705237515225335,
|
| 371 |
+
"grad_norm": 0.35121622681617737,
|
| 372 |
+
"learning_rate": 9.611645588145272e-05,
|
| 373 |
+
"loss": 0.6979084968566894,
|
| 374 |
+
"memory(GiB)": 383.55,
|
| 375 |
+
"step": 175,
|
| 376 |
+
"token_acc": 0.784981684981685,
|
| 377 |
+
"train_speed(iter/s)": 0.055902
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.1753958587088916,
|
| 381 |
+
"grad_norm": 0.41958293318748474,
|
| 382 |
+
"learning_rate": 9.579888921242439e-05,
|
| 383 |
+
"loss": 0.6360678195953369,
|
| 384 |
+
"memory(GiB)": 383.55,
|
| 385 |
+
"step": 180,
|
| 386 |
+
"token_acc": 0.8010867455850961,
|
| 387 |
+
"train_speed(iter/s)": 0.05597
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.18026796589524968,
|
| 391 |
+
"grad_norm": 0.4557216763496399,
|
| 392 |
+
"learning_rate": 9.546941101823963e-05,
|
| 393 |
+
"loss": 0.7268210411071777,
|
| 394 |
+
"memory(GiB)": 383.55,
|
| 395 |
+
"step": 185,
|
| 396 |
+
"token_acc": 0.780511811023622,
|
| 397 |
+
"train_speed(iter/s)": 0.056032
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 0.1851400730816078,
|
| 401 |
+
"grad_norm": 0.44287464022636414,
|
| 402 |
+
"learning_rate": 9.512810699066667e-05,
|
| 403 |
+
"loss": 0.6450634479522706,
|
| 404 |
+
"memory(GiB)": 383.55,
|
| 405 |
+
"step": 190,
|
| 406 |
+
"token_acc": 0.8003896563939072,
|
| 407 |
+
"train_speed(iter/s)": 0.056127
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.1900121802679659,
|
| 411 |
+
"grad_norm": 0.2944161593914032,
|
| 412 |
+
"learning_rate": 9.477506589717518e-05,
|
| 413 |
+
"loss": 0.5534649848937988,
|
| 414 |
+
"memory(GiB)": 383.55,
|
| 415 |
+
"step": 195,
|
| 416 |
+
"token_acc": 0.8173982442138866,
|
| 417 |
+
"train_speed(iter/s)": 0.056116
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.19488428745432398,
|
| 421 |
+
"grad_norm": 0.3743175268173218,
|
| 422 |
+
"learning_rate": 9.441037955784944e-05,
|
| 423 |
+
"loss": 0.7282295227050781,
|
| 424 |
+
"memory(GiB)": 383.55,
|
| 425 |
+
"step": 200,
|
| 426 |
+
"token_acc": 0.7695568400770713,
|
| 427 |
+
"train_speed(iter/s)": 0.05619
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 0.19488428745432398,
|
| 431 |
+
"eval_loss": 0.6866188049316406,
|
| 432 |
+
"eval_runtime": 6.0764,
|
| 433 |
+
"eval_samples_per_second": 0.658,
|
| 434 |
+
"eval_steps_per_second": 0.658,
|
| 435 |
+
"step": 200
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 0.1997563946406821,
|
| 439 |
+
"grad_norm": 0.397987425327301,
|
| 440 |
+
"learning_rate": 9.403414282150738e-05,
|
| 441 |
+
"loss": 0.6911158561706543,
|
| 442 |
+
"memory(GiB)": 383.55,
|
| 443 |
+
"step": 205,
|
| 444 |
+
"token_acc": 0.7807744462859726,
|
| 445 |
+
"train_speed(iter/s)": 0.055478
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 0.2046285018270402,
|
| 449 |
+
"grad_norm": 0.852430522441864,
|
| 450 |
+
"learning_rate": 9.364645354103206e-05,
|
| 451 |
+
"loss": 0.7203257560729981,
|
| 452 |
+
"memory(GiB)": 383.55,
|
| 453 |
+
"step": 210,
|
| 454 |
+
"token_acc": 0.7839163822525598,
|
| 455 |
+
"train_speed(iter/s)": 0.055538
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.2095006090133983,
|
| 459 |
+
"grad_norm": 0.36566832661628723,
|
| 460 |
+
"learning_rate": 9.324741254792171e-05,
|
| 461 |
+
"loss": 0.5751584529876709,
|
| 462 |
+
"memory(GiB)": 383.55,
|
| 463 |
+
"step": 215,
|
| 464 |
+
"token_acc": 0.8120177310786406,
|
| 465 |
+
"train_speed(iter/s)": 0.055665
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.2143727161997564,
|
| 469 |
+
"grad_norm": 0.5105158090591431,
|
| 470 |
+
"learning_rate": 9.28371236260652e-05,
|
| 471 |
+
"loss": 0.5958977699279785,
|
| 472 |
+
"memory(GiB)": 383.55,
|
| 473 |
+
"step": 220,
|
| 474 |
+
"token_acc": 0.8228462471832206,
|
| 475 |
+
"train_speed(iter/s)": 0.055807
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 0.2192448233861145,
|
| 479 |
+
"grad_norm": 0.48028162121772766,
|
| 480 |
+
"learning_rate": 9.241569348474954e-05,
|
| 481 |
+
"loss": 0.7106984615325928,
|
| 482 |
+
"memory(GiB)": 383.55,
|
| 483 |
+
"step": 225,
|
| 484 |
+
"token_acc": 0.7828740844087897,
|
| 485 |
+
"train_speed(iter/s)": 0.055858
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.2241169305724726,
|
| 489 |
+
"grad_norm": 0.32592424750328064,
|
| 490 |
+
"learning_rate": 9.198323173090663e-05,
|
| 491 |
+
"loss": 0.5898131847381591,
|
| 492 |
+
"memory(GiB)": 383.55,
|
| 493 |
+
"step": 230,
|
| 494 |
+
"token_acc": 0.8189450340567084,
|
| 495 |
+
"train_speed(iter/s)": 0.055995
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.2289890377588307,
|
| 499 |
+
"grad_norm": 0.44794151186943054,
|
| 500 |
+
"learning_rate": 9.153985084060623e-05,
|
| 501 |
+
"loss": 0.6042355060577392,
|
| 502 |
+
"memory(GiB)": 383.55,
|
| 503 |
+
"step": 235,
|
| 504 |
+
"token_acc": 0.810122224134963,
|
| 505 |
+
"train_speed(iter/s)": 0.056057
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.23386114494518878,
|
| 509 |
+
"grad_norm": 0.3204025328159332,
|
| 510 |
+
"learning_rate": 9.108566612980298e-05,
|
| 511 |
+
"loss": 0.5558523654937744,
|
| 512 |
+
"memory(GiB)": 383.55,
|
| 513 |
+
"step": 240,
|
| 514 |
+
"token_acc": 0.8260869565217391,
|
| 515 |
+
"train_speed(iter/s)": 0.056072
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.2387332521315469,
|
| 519 |
+
"grad_norm": 0.31540507078170776,
|
| 520 |
+
"learning_rate": 9.062079572434448e-05,
|
| 521 |
+
"loss": 0.6237210273742676,
|
| 522 |
+
"memory(GiB)": 383.55,
|
| 523 |
+
"step": 245,
|
| 524 |
+
"token_acc": 0.8021445866482186,
|
| 525 |
+
"train_speed(iter/s)": 0.056086
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.243605359317905,
|
| 529 |
+
"grad_norm": 0.619088888168335,
|
| 530 |
+
"learning_rate": 9.014536052924883e-05,
|
| 531 |
+
"loss": 0.664583158493042,
|
| 532 |
+
"memory(GiB)": 383.55,
|
| 533 |
+
"step": 250,
|
| 534 |
+
"token_acc": 0.793002915451895,
|
| 535 |
+
"train_speed(iter/s)": 0.056116
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.24847746650426308,
|
| 539 |
+
"grad_norm": 0.6715230345726013,
|
| 540 |
+
"learning_rate": 8.965948419725922e-05,
|
| 541 |
+
"loss": 0.5711063861846923,
|
| 542 |
+
"memory(GiB)": 383.55,
|
| 543 |
+
"step": 255,
|
| 544 |
+
"token_acc": 0.8185620394343757,
|
| 545 |
+
"train_speed(iter/s)": 0.055175
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 0.25334957369062117,
|
| 549 |
+
"grad_norm": 0.4514237642288208,
|
| 550 |
+
"learning_rate": 8.916329309668397e-05,
|
| 551 |
+
"loss": 0.721324348449707,
|
| 552 |
+
"memory(GiB)": 383.55,
|
| 553 |
+
"step": 260,
|
| 554 |
+
"token_acc": 0.7792865828942035,
|
| 555 |
+
"train_speed(iter/s)": 0.055266
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.2582216808769793,
|
| 559 |
+
"grad_norm": 0.5026947855949402,
|
| 560 |
+
"learning_rate": 8.865691627853013e-05,
|
| 561 |
+
"loss": 0.6661148548126221,
|
| 562 |
+
"memory(GiB)": 383.55,
|
| 563 |
+
"step": 265,
|
| 564 |
+
"token_acc": 0.7951268025857782,
|
| 565 |
+
"train_speed(iter/s)": 0.05533
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 0.2630937880633374,
|
| 569 |
+
"grad_norm": 0.3138331174850464,
|
| 570 |
+
"learning_rate": 8.814048544293965e-05,
|
| 571 |
+
"loss": 0.6717385292053223,
|
| 572 |
+
"memory(GiB)": 383.55,
|
| 573 |
+
"step": 270,
|
| 574 |
+
"token_acc": 0.7904462355022607,
|
| 575 |
+
"train_speed(iter/s)": 0.055296
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 0.2679658952496955,
|
| 579 |
+
"grad_norm": 0.3270625174045563,
|
| 580 |
+
"learning_rate": 8.76141349049362e-05,
|
| 581 |
+
"loss": 0.6027359008789063,
|
| 582 |
+
"memory(GiB)": 383.55,
|
| 583 |
+
"step": 275,
|
| 584 |
+
"token_acc": 0.8082101806239738,
|
| 585 |
+
"train_speed(iter/s)": 0.05525
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.2728380024360536,
|
| 589 |
+
"grad_norm": 0.4341810941696167,
|
| 590 |
+
"learning_rate": 8.707800155949217e-05,
|
| 591 |
+
"loss": 0.6553579330444336,
|
| 592 |
+
"memory(GiB)": 383.55,
|
| 593 |
+
"step": 280,
|
| 594 |
+
"token_acc": 0.797032640949555,
|
| 595 |
+
"train_speed(iter/s)": 0.055271
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 0.2777101096224117,
|
| 599 |
+
"grad_norm": 0.37805306911468506,
|
| 600 |
+
"learning_rate": 8.653222484592458e-05,
|
| 601 |
+
"loss": 0.6515018463134765,
|
| 602 |
+
"memory(GiB)": 383.55,
|
| 603 |
+
"step": 285,
|
| 604 |
+
"token_acc": 0.794751477233229,
|
| 605 |
+
"train_speed(iter/s)": 0.05524
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.28258221680876977,
|
| 609 |
+
"grad_norm": 0.38902854919433594,
|
| 610 |
+
"learning_rate": 8.597694671162921e-05,
|
| 611 |
+
"loss": 0.592349624633789,
|
| 612 |
+
"memory(GiB)": 383.55,
|
| 613 |
+
"step": 290,
|
| 614 |
+
"token_acc": 0.815828677839851,
|
| 615 |
+
"train_speed(iter/s)": 0.05519
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 0.2874543239951279,
|
| 619 |
+
"grad_norm": 0.3007030487060547,
|
| 620 |
+
"learning_rate": 8.541231157516247e-05,
|
| 621 |
+
"loss": 0.6616343021392822,
|
| 622 |
+
"memory(GiB)": 383.55,
|
| 623 |
+
"step": 295,
|
| 624 |
+
"token_acc": 0.7961879284400601,
|
| 625 |
+
"train_speed(iter/s)": 0.05519
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.292326431181486,
|
| 629 |
+
"grad_norm": 0.43431806564331055,
|
| 630 |
+
"learning_rate": 8.483846628868055e-05,
|
| 631 |
+
"loss": 0.6408910751342773,
|
| 632 |
+
"memory(GiB)": 383.55,
|
| 633 |
+
"step": 300,
|
| 634 |
+
"token_acc": 0.7999295526593871,
|
| 635 |
+
"train_speed(iter/s)": 0.055286
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 0.292326431181486,
|
| 639 |
+
"eval_loss": 0.6525390148162842,
|
| 640 |
+
"eval_runtime": 6.2823,
|
| 641 |
+
"eval_samples_per_second": 0.637,
|
| 642 |
+
"eval_steps_per_second": 0.637,
|
| 643 |
+
"step": 300
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 0.2971985383678441,
|
| 647 |
+
"grad_norm": 0.4932222068309784,
|
| 648 |
+
"learning_rate": 8.425556009974566e-05,
|
| 649 |
+
"loss": 0.6335715770721435,
|
| 650 |
+
"memory(GiB)": 383.55,
|
| 651 |
+
"step": 305,
|
| 652 |
+
"token_acc": 0.8036400066789113,
|
| 653 |
+
"train_speed(iter/s)": 0.054583
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.3020706455542022,
|
| 657 |
+
"grad_norm": 0.27842646837234497,
|
| 658 |
+
"learning_rate": 8.366374461250916e-05,
|
| 659 |
+
"loss": 0.570946216583252,
|
| 660 |
+
"memory(GiB)": 383.55,
|
| 661 |
+
"step": 310,
|
| 662 |
+
"token_acc": 0.8238350381555447,
|
| 663 |
+
"train_speed(iter/s)": 0.054659
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 0.30694275274056027,
|
| 667 |
+
"grad_norm": 0.7104659080505371,
|
| 668 |
+
"learning_rate": 8.306317374828194e-05,
|
| 669 |
+
"loss": 0.566010570526123,
|
| 670 |
+
"memory(GiB)": 383.55,
|
| 671 |
+
"step": 315,
|
| 672 |
+
"token_acc": 0.8189669219488349,
|
| 673 |
+
"train_speed(iter/s)": 0.054662
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 0.3118148599269184,
|
| 677 |
+
"grad_norm": 0.8684744834899902,
|
| 678 |
+
"learning_rate": 8.245400370550198e-05,
|
| 679 |
+
"loss": 0.677960729598999,
|
| 680 |
+
"memory(GiB)": 383.55,
|
| 681 |
+
"step": 320,
|
| 682 |
+
"token_acc": 0.7772163527790538,
|
| 683 |
+
"train_speed(iter/s)": 0.054794
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.3166869671132765,
|
| 687 |
+
"grad_norm": 0.3846539258956909,
|
| 688 |
+
"learning_rate": 8.183639291910987e-05,
|
| 689 |
+
"loss": 0.5622167587280273,
|
| 690 |
+
"memory(GiB)": 383.55,
|
| 691 |
+
"step": 325,
|
| 692 |
+
"token_acc": 0.830480089318943,
|
| 693 |
+
"train_speed(iter/s)": 0.054821
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 0.3215590742996346,
|
| 697 |
+
"grad_norm": 0.34010785818099976,
|
| 698 |
+
"learning_rate": 8.121050201934235e-05,
|
| 699 |
+
"loss": 0.5877705574035644,
|
| 700 |
+
"memory(GiB)": 383.55,
|
| 701 |
+
"step": 330,
|
| 702 |
+
"token_acc": 0.8287964389659305,
|
| 703 |
+
"train_speed(iter/s)": 0.054899
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.3264311814859927,
|
| 707 |
+
"grad_norm": 0.3751339912414551,
|
| 708 |
+
"learning_rate": 8.057649378995526e-05,
|
| 709 |
+
"loss": 0.5179604053497314,
|
| 710 |
+
"memory(GiB)": 383.55,
|
| 711 |
+
"step": 335,
|
| 712 |
+
"token_acc": 0.8402439024390244,
|
| 713 |
+
"train_speed(iter/s)": 0.054839
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.3313032886723508,
|
| 717 |
+
"grad_norm": 0.3137739896774292,
|
| 718 |
+
"learning_rate": 7.993453312588607e-05,
|
| 719 |
+
"loss": 0.5339327335357666,
|
| 720 |
+
"memory(GiB)": 383.55,
|
| 721 |
+
"step": 340,
|
| 722 |
+
"token_acc": 0.8365357839042049,
|
| 723 |
+
"train_speed(iter/s)": 0.054764
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.33617539585870887,
|
| 727 |
+
"grad_norm": 0.5746834874153137,
|
| 728 |
+
"learning_rate": 7.928478699036755e-05,
|
| 729 |
+
"loss": 0.6346034049987793,
|
| 730 |
+
"memory(GiB)": 383.55,
|
| 731 |
+
"step": 345,
|
| 732 |
+
"token_acc": 0.7972016183412003,
|
| 733 |
+
"train_speed(iter/s)": 0.054713
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 0.341047503045067,
|
| 737 |
+
"grad_norm": 0.3580325245857239,
|
| 738 |
+
"learning_rate": 7.862742437150336e-05,
|
| 739 |
+
"loss": 0.6608481884002686,
|
| 740 |
+
"memory(GiB)": 383.55,
|
| 741 |
+
"step": 350,
|
| 742 |
+
"token_acc": 0.7929736511919699,
|
| 743 |
+
"train_speed(iter/s)": 0.054762
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 0.3459196102314251,
|
| 747 |
+
"grad_norm": 0.4622519612312317,
|
| 748 |
+
"learning_rate": 7.796261623831713e-05,
|
| 749 |
+
"loss": 0.562419080734253,
|
| 750 |
+
"memory(GiB)": 383.55,
|
| 751 |
+
"step": 355,
|
| 752 |
+
"token_acc": 0.8190336211647988,
|
| 753 |
+
"train_speed(iter/s)": 0.054396
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.3507917174177832,
|
| 757 |
+
"grad_norm": 0.5616739392280579,
|
| 758 |
+
"learning_rate": 7.729053549628622e-05,
|
| 759 |
+
"loss": 0.5495719909667969,
|
| 760 |
+
"memory(GiB)": 383.55,
|
| 761 |
+
"step": 360,
|
| 762 |
+
"token_acc": 0.8339377743844245,
|
| 763 |
+
"train_speed(iter/s)": 0.054442
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 0.3556638246041413,
|
| 767 |
+
"grad_norm": 0.7364129424095154,
|
| 768 |
+
"learning_rate": 7.661135694237198e-05,
|
| 769 |
+
"loss": 0.4548810958862305,
|
| 770 |
+
"memory(GiB)": 387.42,
|
| 771 |
+
"step": 365,
|
| 772 |
+
"token_acc": 0.8370827285921626,
|
| 773 |
+
"train_speed(iter/s)": 0.054383
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.36053593179049936,
|
| 777 |
+
"grad_norm": 0.44831952452659607,
|
| 778 |
+
"learning_rate": 7.592525721955786e-05,
|
| 779 |
+
"loss": 0.5882142066955567,
|
| 780 |
+
"memory(GiB)": 387.42,
|
| 781 |
+
"step": 370,
|
| 782 |
+
"token_acc": 0.8161894662424886,
|
| 783 |
+
"train_speed(iter/s)": 0.054337
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.3654080389768575,
|
| 787 |
+
"grad_norm": 0.37750759720802307,
|
| 788 |
+
"learning_rate": 7.523241477090763e-05,
|
| 789 |
+
"loss": 0.6884512901306152,
|
| 790 |
+
"memory(GiB)": 387.42,
|
| 791 |
+
"step": 375,
|
| 792 |
+
"token_acc": 0.7952127659574468,
|
| 793 |
+
"train_speed(iter/s)": 0.054385
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.3702801461632156,
|
| 797 |
+
"grad_norm": 0.5074845552444458,
|
| 798 |
+
"learning_rate": 7.45330097931553e-05,
|
| 799 |
+
"loss": 0.5458427906036377,
|
| 800 |
+
"memory(GiB)": 387.42,
|
| 801 |
+
"step": 380,
|
| 802 |
+
"token_acc": 0.8217197924388436,
|
| 803 |
+
"train_speed(iter/s)": 0.054354
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 0.3751522533495737,
|
| 807 |
+
"grad_norm": 0.6083484292030334,
|
| 808 |
+
"learning_rate": 7.382722418983892e-05,
|
| 809 |
+
"loss": 0.5680232048034668,
|
| 810 |
+
"memory(GiB)": 387.42,
|
| 811 |
+
"step": 385,
|
| 812 |
+
"token_acc": 0.8248374239563667,
|
| 813 |
+
"train_speed(iter/s)": 0.054329
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 0.3800243605359318,
|
| 817 |
+
"grad_norm": 0.39138278365135193,
|
| 818 |
+
"learning_rate": 7.311524152399054e-05,
|
| 819 |
+
"loss": 0.7077183246612548,
|
| 820 |
+
"memory(GiB)": 387.42,
|
| 821 |
+
"step": 390,
|
| 822 |
+
"token_acc": 0.7912014292094686,
|
| 823 |
+
"train_speed(iter/s)": 0.054329
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.3848964677222899,
|
| 827 |
+
"grad_norm": 0.4244479238986969,
|
| 828 |
+
"learning_rate": 7.239724697039457e-05,
|
| 829 |
+
"loss": 0.6999778270721435,
|
| 830 |
+
"memory(GiB)": 387.42,
|
| 831 |
+
"step": 395,
|
| 832 |
+
"token_acc": 0.7828650029475339,
|
| 833 |
+
"train_speed(iter/s)": 0.054413
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 0.38976857490864797,
|
| 837 |
+
"grad_norm": 0.3658107817173004,
|
| 838 |
+
"learning_rate": 7.167342726742685e-05,
|
| 839 |
+
"loss": 0.5321448802947998,
|
| 840 |
+
"memory(GiB)": 387.42,
|
| 841 |
+
"step": 400,
|
| 842 |
+
"token_acc": 0.8257604205782951,
|
| 843 |
+
"train_speed(iter/s)": 0.054414
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.38976857490864797,
|
| 847 |
+
"eval_loss": 0.647614598274231,
|
| 848 |
+
"eval_runtime": 6.1299,
|
| 849 |
+
"eval_samples_per_second": 0.653,
|
| 850 |
+
"eval_steps_per_second": 0.653,
|
| 851 |
+
"step": 400
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 0.3946406820950061,
|
| 855 |
+
"grad_norm": 0.4579378068447113,
|
| 856 |
+
"learning_rate": 7.094397066848716e-05,
|
| 857 |
+
"loss": 0.6339591979980469,
|
| 858 |
+
"memory(GiB)": 387.42,
|
| 859 |
+
"step": 405,
|
| 860 |
+
"token_acc": 0.7953757225433526,
|
| 861 |
+
"train_speed(iter/s)": 0.054198
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 0.3995127892813642,
|
| 865 |
+
"grad_norm": 0.41108816862106323,
|
| 866 |
+
"learning_rate": 7.020906689303766e-05,
|
| 867 |
+
"loss": 0.6498037338256836,
|
| 868 |
+
"memory(GiB)": 387.42,
|
| 869 |
+
"step": 410,
|
| 870 |
+
"token_acc": 0.8013311819281969,
|
| 871 |
+
"train_speed(iter/s)": 0.054274
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.4043848964677223,
|
| 875 |
+
"grad_norm": 0.3730790615081787,
|
| 876 |
+
"learning_rate": 6.946890707726004e-05,
|
| 877 |
+
"loss": 0.6224189281463623,
|
| 878 |
+
"memory(GiB)": 387.42,
|
| 879 |
+
"step": 415,
|
| 880 |
+
"token_acc": 0.8109767441860465,
|
| 881 |
+
"train_speed(iter/s)": 0.054342
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 0.4092570036540804,
|
| 885 |
+
"grad_norm": 0.41862693428993225,
|
| 886 |
+
"learning_rate": 6.872368372434416e-05,
|
| 887 |
+
"loss": 0.6285569190979003,
|
| 888 |
+
"memory(GiB)": 387.42,
|
| 889 |
+
"step": 420,
|
| 890 |
+
"token_acc": 0.793915399041467,
|
| 891 |
+
"train_speed(iter/s)": 0.054335
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.41412911084043846,
|
| 895 |
+
"grad_norm": 0.4861293435096741,
|
| 896 |
+
"learning_rate": 6.797359065442117e-05,
|
| 897 |
+
"loss": 0.5771468162536622,
|
| 898 |
+
"memory(GiB)": 387.42,
|
| 899 |
+
"step": 425,
|
| 900 |
+
"token_acc": 0.8196579720158922,
|
| 901 |
+
"train_speed(iter/s)": 0.054339
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 0.4190012180267966,
|
| 905 |
+
"grad_norm": 0.30941805243492126,
|
| 906 |
+
"learning_rate": 6.721882295415425e-05,
|
| 907 |
+
"loss": 0.5844586372375489,
|
| 908 |
+
"memory(GiB)": 387.42,
|
| 909 |
+
"step": 430,
|
| 910 |
+
"token_acc": 0.814694173000362,
|
| 911 |
+
"train_speed(iter/s)": 0.05432
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.4238733252131547,
|
| 915 |
+
"grad_norm": 0.3820112645626068,
|
| 916 |
+
"learning_rate": 6.645957692599969e-05,
|
| 917 |
+
"loss": 0.5823289394378662,
|
| 918 |
+
"memory(GiB)": 387.42,
|
| 919 |
+
"step": 435,
|
| 920 |
+
"token_acc": 0.8027565654684299,
|
| 921 |
+
"train_speed(iter/s)": 0.054382
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.4287454323995128,
|
| 925 |
+
"grad_norm": 0.3910198509693146,
|
| 926 |
+
"learning_rate": 6.569605003715201e-05,
|
| 927 |
+
"loss": 0.561509084701538,
|
| 928 |
+
"memory(GiB)": 387.42,
|
| 929 |
+
"step": 440,
|
| 930 |
+
"token_acc": 0.8264751552795031,
|
| 931 |
+
"train_speed(iter/s)": 0.054462
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 0.4336175395858709,
|
| 935 |
+
"grad_norm": 0.3805302381515503,
|
| 936 |
+
"learning_rate": 6.492844086818599e-05,
|
| 937 |
+
"loss": 0.558375883102417,
|
| 938 |
+
"memory(GiB)": 387.42,
|
| 939 |
+
"step": 445,
|
| 940 |
+
"token_acc": 0.8262056414922657,
|
| 941 |
+
"train_speed(iter/s)": 0.05444
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.438489646772229,
|
| 945 |
+
"grad_norm": 0.6036235690116882,
|
| 946 |
+
"learning_rate": 6.41569490614092e-05,
|
| 947 |
+
"loss": 0.6268420696258545,
|
| 948 |
+
"memory(GiB)": 387.42,
|
| 949 |
+
"step": 450,
|
| 950 |
+
"token_acc": 0.8061224489795918,
|
| 951 |
+
"train_speed(iter/s)": 0.054446
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 0.44336175395858707,
|
| 955 |
+
"grad_norm": 0.4275857210159302,
|
| 956 |
+
"learning_rate": 6.338177526893836e-05,
|
| 957 |
+
"loss": 0.5441042423248291,
|
| 958 |
+
"memory(GiB)": 387.42,
|
| 959 |
+
"step": 455,
|
| 960 |
+
"token_acc": 0.8360881542699724,
|
| 961 |
+
"train_speed(iter/s)": 0.05418
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.4482338611449452,
|
| 965 |
+
"grad_norm": 0.4830683469772339,
|
| 966 |
+
"learning_rate": 6.260312110051312e-05,
|
| 967 |
+
"loss": 0.606513261795044,
|
| 968 |
+
"memory(GiB)": 387.42,
|
| 969 |
+
"step": 460,
|
| 970 |
+
"token_acc": 0.8049238864875023,
|
| 971 |
+
"train_speed(iter/s)": 0.054224
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.4531059683313033,
|
| 975 |
+
"grad_norm": 0.35629284381866455,
|
| 976 |
+
"learning_rate": 6.182118907106068e-05,
|
| 977 |
+
"loss": 0.538546371459961,
|
| 978 |
+
"memory(GiB)": 387.42,
|
| 979 |
+
"step": 465,
|
| 980 |
+
"token_acc": 0.8373831775700935,
|
| 981 |
+
"train_speed(iter/s)": 0.054204
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 0.4579780755176614,
|
| 985 |
+
"grad_norm": 0.46749940514564514,
|
| 986 |
+
"learning_rate": 6.103618254802511e-05,
|
| 987 |
+
"loss": 0.5923898696899415,
|
| 988 |
+
"memory(GiB)": 387.42,
|
| 989 |
+
"step": 470,
|
| 990 |
+
"token_acc": 0.8042936553574851,
|
| 991 |
+
"train_speed(iter/s)": 0.054261
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 0.4628501827040195,
|
| 995 |
+
"grad_norm": 0.6278035044670105,
|
| 996 |
+
"learning_rate": 6.024830569847477e-05,
|
| 997 |
+
"loss": 0.5971939086914062,
|
| 998 |
+
"memory(GiB)": 387.42,
|
| 999 |
+
"step": 475,
|
| 1000 |
+
"token_acc": 0.8176121372031663,
|
| 1001 |
+
"train_speed(iter/s)": 0.054245
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 0.46772228989037756,
|
| 1005 |
+
"grad_norm": 0.3572694957256317,
|
| 1006 |
+
"learning_rate": 5.945776343600207e-05,
|
| 1007 |
+
"loss": 0.5843085765838623,
|
| 1008 |
+
"memory(GiB)": 387.42,
|
| 1009 |
+
"step": 480,
|
| 1010 |
+
"token_acc": 0.8212882953652789,
|
| 1011 |
+
"train_speed(iter/s)": 0.054246
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.4725943970767357,
|
| 1015 |
+
"grad_norm": 0.5189170241355896,
|
| 1016 |
+
"learning_rate": 5.866476136742862e-05,
|
| 1017 |
+
"loss": 0.5234210968017579,
|
| 1018 |
+
"memory(GiB)": 387.42,
|
| 1019 |
+
"step": 485,
|
| 1020 |
+
"token_acc": 0.8463819691577699,
|
| 1021 |
+
"train_speed(iter/s)": 0.05426
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 0.4774665042630938,
|
| 1025 |
+
"grad_norm": 0.41832658648490906,
|
| 1026 |
+
"learning_rate": 5.7869505739330546e-05,
|
| 1027 |
+
"loss": 0.6695927619934082,
|
| 1028 |
+
"memory(GiB)": 387.42,
|
| 1029 |
+
"step": 490,
|
| 1030 |
+
"token_acc": 0.7924812030075188,
|
| 1031 |
+
"train_speed(iter/s)": 0.05433
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.4823386114494519,
|
| 1035 |
+
"grad_norm": 4.011805534362793,
|
| 1036 |
+
"learning_rate": 5.7072203384397064e-05,
|
| 1037 |
+
"loss": 0.5814547538757324,
|
| 1038 |
+
"memory(GiB)": 387.42,
|
| 1039 |
+
"step": 495,
|
| 1040 |
+
"token_acc": 0.8110627719080175,
|
| 1041 |
+
"train_speed(iter/s)": 0.054376
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.48721071863581,
|
| 1045 |
+
"grad_norm": 0.31671130657196045,
|
| 1046 |
+
"learning_rate": 5.627306166763684e-05,
|
| 1047 |
+
"loss": 0.5855265617370605,
|
| 1048 |
+
"memory(GiB)": 387.42,
|
| 1049 |
+
"step": 500,
|
| 1050 |
+
"token_acc": 0.8094142629623076,
|
| 1051 |
+
"train_speed(iter/s)": 0.054362
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 0.48721071863581,
|
| 1055 |
+
"eval_loss": 0.6302051544189453,
|
| 1056 |
+
"eval_runtime": 6.1545,
|
| 1057 |
+
"eval_samples_per_second": 0.65,
|
| 1058 |
+
"eval_steps_per_second": 0.65,
|
| 1059 |
+
"step": 500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.4920828258221681,
|
| 1063 |
+
"grad_norm": 0.3875284194946289,
|
| 1064 |
+
"learning_rate": 5.5472288432445774e-05,
|
| 1065 |
+
"loss": 0.59937744140625,
|
| 1066 |
+
"memory(GiB)": 387.42,
|
| 1067 |
+
"step": 505,
|
| 1068 |
+
"token_acc": 0.7988918837975442,
|
| 1069 |
+
"train_speed(iter/s)": 0.05424
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 0.49695493300852617,
|
| 1073 |
+
"grad_norm": 0.4411413371562958,
|
| 1074 |
+
"learning_rate": 5.467009194655045e-05,
|
| 1075 |
+
"loss": 0.5820174217224121,
|
| 1076 |
+
"memory(GiB)": 387.42,
|
| 1077 |
+
"step": 510,
|
| 1078 |
+
"token_acc": 0.8234812510234157,
|
| 1079 |
+
"train_speed(iter/s)": 0.054197
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 0.5018270401948843,
|
| 1083 |
+
"grad_norm": 0.5111451148986816,
|
| 1084 |
+
"learning_rate": 5.386668084784112e-05,
|
| 1085 |
+
"loss": 0.5154130935668946,
|
| 1086 |
+
"memory(GiB)": 387.42,
|
| 1087 |
+
"step": 515,
|
| 1088 |
+
"token_acc": 0.8397686998694274,
|
| 1089 |
+
"train_speed(iter/s)": 0.05426
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.5066991473812423,
|
| 1093 |
+
"grad_norm": 0.29832109808921814,
|
| 1094 |
+
"learning_rate": 5.306226409010855e-05,
|
| 1095 |
+
"loss": 0.5672587394714356,
|
| 1096 |
+
"memory(GiB)": 387.42,
|
| 1097 |
+
"step": 520,
|
| 1098 |
+
"token_acc": 0.8263521756811713,
|
| 1099 |
+
"train_speed(iter/s)": 0.054274
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 0.5115712545676004,
|
| 1103 |
+
"grad_norm": 0.42139527201652527,
|
| 1104 |
+
"learning_rate": 5.22570508886986e-05,
|
| 1105 |
+
"loss": 0.5327470302581787,
|
| 1106 |
+
"memory(GiB)": 387.42,
|
| 1107 |
+
"step": 525,
|
| 1108 |
+
"token_acc": 0.8310478199718706,
|
| 1109 |
+
"train_speed(iter/s)": 0.054332
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.5164433617539586,
|
| 1113 |
+
"grad_norm": 0.34750285744667053,
|
| 1114 |
+
"learning_rate": 5.145125066609877e-05,
|
| 1115 |
+
"loss": 0.61210618019104,
|
| 1116 |
+
"memory(GiB)": 387.42,
|
| 1117 |
+
"step": 530,
|
| 1118 |
+
"token_acc": 0.8104413702239789,
|
| 1119 |
+
"train_speed(iter/s)": 0.054325
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.5213154689403167,
|
| 1123 |
+
"grad_norm": 0.5557289123535156,
|
| 1124 |
+
"learning_rate": 5.0645072997471e-05,
|
| 1125 |
+
"loss": 0.5486731052398681,
|
| 1126 |
+
"memory(GiB)": 387.42,
|
| 1127 |
+
"step": 535,
|
| 1128 |
+
"token_acc": 0.8223992502343018,
|
| 1129 |
+
"train_speed(iter/s)": 0.054295
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.5261875761266748,
|
| 1133 |
+
"grad_norm": 1.370209813117981,
|
| 1134 |
+
"learning_rate": 4.983872755614461e-05,
|
| 1135 |
+
"loss": 0.6499679565429688,
|
| 1136 |
+
"memory(GiB)": 387.42,
|
| 1137 |
+
"step": 540,
|
| 1138 |
+
"token_acc": 0.7975866095757104,
|
| 1139 |
+
"train_speed(iter/s)": 0.054348
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 0.5310596833130329,
|
| 1143 |
+
"grad_norm": 0.4371365010738373,
|
| 1144 |
+
"learning_rate": 4.9032424059083774e-05,
|
| 1145 |
+
"loss": 0.43409147262573244,
|
| 1146 |
+
"memory(GiB)": 387.42,
|
| 1147 |
+
"step": 545,
|
| 1148 |
+
"token_acc": 0.8684942391736193,
|
| 1149 |
+
"train_speed(iter/s)": 0.054321
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 0.535931790499391,
|
| 1153 |
+
"grad_norm": 0.4735865890979767,
|
| 1154 |
+
"learning_rate": 4.8226372212343726e-05,
|
| 1155 |
+
"loss": 0.5776564598083496,
|
| 1156 |
+
"memory(GiB)": 387.42,
|
| 1157 |
+
"step": 550,
|
| 1158 |
+
"token_acc": 0.8255653883972468,
|
| 1159 |
+
"train_speed(iter/s)": 0.054368
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 0.5408038976857491,
|
| 1163 |
+
"grad_norm": 0.6005700826644897,
|
| 1164 |
+
"learning_rate": 4.742078165652958e-05,
|
| 1165 |
+
"loss": 0.5744057178497315,
|
| 1166 |
+
"memory(GiB)": 387.42,
|
| 1167 |
+
"step": 555,
|
| 1168 |
+
"token_acc": 0.8105436573311368,
|
| 1169 |
+
"train_speed(iter/s)": 0.054325
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 0.5456760048721072,
|
| 1173 |
+
"grad_norm": 0.4128513038158417,
|
| 1174 |
+
"learning_rate": 4.661586191227247e-05,
|
| 1175 |
+
"loss": 0.5321125030517578,
|
| 1176 |
+
"memory(GiB)": 387.42,
|
| 1177 |
+
"step": 560,
|
| 1178 |
+
"token_acc": 0.8245080500894454,
|
| 1179 |
+
"train_speed(iter/s)": 0.054305
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.5505481120584653,
|
| 1183 |
+
"grad_norm": 0.4688722491264343,
|
| 1184 |
+
"learning_rate": 4.581182232573658e-05,
|
| 1185 |
+
"loss": 0.5235236167907715,
|
| 1186 |
+
"memory(GiB)": 387.42,
|
| 1187 |
+
"step": 565,
|
| 1188 |
+
"token_acc": 0.8205183122724352,
|
| 1189 |
+
"train_speed(iter/s)": 0.054352
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.5554202192448234,
|
| 1193 |
+
"grad_norm": 0.4604549705982208,
|
| 1194 |
+
"learning_rate": 4.500887201417187e-05,
|
| 1195 |
+
"loss": 0.6571295261383057,
|
| 1196 |
+
"memory(GiB)": 387.42,
|
| 1197 |
+
"step": 570,
|
| 1198 |
+
"token_acc": 0.8019607843137255,
|
| 1199 |
+
"train_speed(iter/s)": 0.054361
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.5602923264311814,
|
| 1203 |
+
"grad_norm": 0.48336780071258545,
|
| 1204 |
+
"learning_rate": 4.4207219811526056e-05,
|
| 1205 |
+
"loss": 0.5963138580322266,
|
| 1206 |
+
"memory(GiB)": 387.42,
|
| 1207 |
+
"step": 575,
|
| 1208 |
+
"token_acc": 0.8077416987708678,
|
| 1209 |
+
"train_speed(iter/s)": 0.054409
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 0.5651644336175395,
|
| 1213 |
+
"grad_norm": 0.5700681805610657,
|
| 1214 |
+
"learning_rate": 4.3407074214130446e-05,
|
| 1215 |
+
"loss": 0.6309503555297852,
|
| 1216 |
+
"memory(GiB)": 387.42,
|
| 1217 |
+
"step": 580,
|
| 1218 |
+
"token_acc": 0.7960770454143842,
|
| 1219 |
+
"train_speed(iter/s)": 0.054412
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 0.5700365408038977,
|
| 1223 |
+
"grad_norm": 0.40493443608283997,
|
| 1224 |
+
"learning_rate": 4.2608643326473496e-05,
|
| 1225 |
+
"loss": 0.5265829563140869,
|
| 1226 |
+
"memory(GiB)": 387.42,
|
| 1227 |
+
"step": 585,
|
| 1228 |
+
"token_acc": 0.8364477970169724,
|
| 1229 |
+
"train_speed(iter/s)": 0.054419
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 0.5749086479902558,
|
| 1233 |
+
"grad_norm": 0.42441654205322266,
|
| 1234 |
+
"learning_rate": 4.181213480707637e-05,
|
| 1235 |
+
"loss": 0.5463868618011475,
|
| 1236 |
+
"memory(GiB)": 387.42,
|
| 1237 |
+
"step": 590,
|
| 1238 |
+
"token_acc": 0.8250831178426302,
|
| 1239 |
+
"train_speed(iter/s)": 0.054415
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 0.5797807551766139,
|
| 1243 |
+
"grad_norm": 0.5273870825767517,
|
| 1244 |
+
"learning_rate": 4.1017755814484374e-05,
|
| 1245 |
+
"loss": 0.6219929218292236,
|
| 1246 |
+
"memory(GiB)": 387.42,
|
| 1247 |
+
"step": 595,
|
| 1248 |
+
"token_acc": 0.8101965601965602,
|
| 1249 |
+
"train_speed(iter/s)": 0.054492
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.584652862362972,
|
| 1253 |
+
"grad_norm": 0.5027340650558472,
|
| 1254 |
+
"learning_rate": 4.0225712953388494e-05,
|
| 1255 |
+
"loss": 0.47921223640441896,
|
| 1256 |
+
"memory(GiB)": 387.42,
|
| 1257 |
+
"step": 600,
|
| 1258 |
+
"token_acc": 0.8507462686567164,
|
| 1259 |
+
"train_speed(iter/s)": 0.054456
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 0.584652862362972,
|
| 1263 |
+
"eval_loss": 0.5931864976882935,
|
| 1264 |
+
"eval_runtime": 6.2202,
|
| 1265 |
+
"eval_samples_per_second": 0.643,
|
| 1266 |
+
"eval_steps_per_second": 0.643,
|
| 1267 |
+
"step": 600
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 0.5895249695493301,
|
| 1271 |
+
"grad_norm": 0.7974056005477905,
|
| 1272 |
+
"learning_rate": 3.943621222089102e-05,
|
| 1273 |
+
"loss": 0.5052922248840332,
|
| 1274 |
+
"memory(GiB)": 387.42,
|
| 1275 |
+
"step": 605,
|
| 1276 |
+
"token_acc": 0.8312937062937062,
|
| 1277 |
+
"train_speed(iter/s)": 0.054258
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.5943970767356882,
|
| 1281 |
+
"grad_norm": 0.38420093059539795,
|
| 1282 |
+
"learning_rate": 3.864945895292908e-05,
|
| 1283 |
+
"loss": 0.5411774635314941,
|
| 1284 |
+
"memory(GiB)": 387.42,
|
| 1285 |
+
"step": 610,
|
| 1286 |
+
"token_acc": 0.8309124767225325,
|
| 1287 |
+
"train_speed(iter/s)": 0.054201
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.5992691839220463,
|
| 1291 |
+
"grad_norm": 0.9411633014678955,
|
| 1292 |
+
"learning_rate": 3.786565777087022e-05,
|
| 1293 |
+
"loss": 0.6929959297180176,
|
| 1294 |
+
"memory(GiB)": 387.42,
|
| 1295 |
+
"step": 615,
|
| 1296 |
+
"token_acc": 0.7847842261904762,
|
| 1297 |
+
"train_speed(iter/s)": 0.05425
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.6041412911084044,
|
| 1301 |
+
"grad_norm": 0.35226595401763916,
|
| 1302 |
+
"learning_rate": 3.708501252829386e-05,
|
| 1303 |
+
"loss": 0.5966301918029785,
|
| 1304 |
+
"memory(GiB)": 387.42,
|
| 1305 |
+
"step": 620,
|
| 1306 |
+
"token_acc": 0.8161076443057722,
|
| 1307 |
+
"train_speed(iter/s)": 0.054233
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 0.6090133982947625,
|
| 1311 |
+
"grad_norm": 0.4208815097808838,
|
| 1312 |
+
"learning_rate": 3.6307726257972255e-05,
|
| 1313 |
+
"loss": 0.5394818782806396,
|
| 1314 |
+
"memory(GiB)": 387.42,
|
| 1315 |
+
"step": 625,
|
| 1316 |
+
"token_acc": 0.8257628294036061,
|
| 1317 |
+
"train_speed(iter/s)": 0.054209
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 0.6138855054811205,
|
| 1321 |
+
"grad_norm": 0.445925772190094,
|
| 1322 |
+
"learning_rate": 3.553400111906523e-05,
|
| 1323 |
+
"loss": 0.6164620399475098,
|
| 1324 |
+
"memory(GiB)": 387.42,
|
| 1325 |
+
"step": 630,
|
| 1326 |
+
"token_acc": 0.8090881366270204,
|
| 1327 |
+
"train_speed(iter/s)": 0.054222
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 0.6187576126674786,
|
| 1331 |
+
"grad_norm": 0.5922476649284363,
|
| 1332 |
+
"learning_rate": 3.476403834454183e-05,
|
| 1333 |
+
"loss": 0.5115623474121094,
|
| 1334 |
+
"memory(GiB)": 387.42,
|
| 1335 |
+
"step": 635,
|
| 1336 |
+
"token_acc": 0.8346325167037862,
|
| 1337 |
+
"train_speed(iter/s)": 0.054244
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 0.6236297198538368,
|
| 1341 |
+
"grad_norm": 0.5026776790618896,
|
| 1342 |
+
"learning_rate": 3.399803818884311e-05,
|
| 1343 |
+
"loss": 0.5328683853149414,
|
| 1344 |
+
"memory(GiB)": 387.42,
|
| 1345 |
+
"step": 640,
|
| 1346 |
+
"token_acc": 0.8462420173571311,
|
| 1347 |
+
"train_speed(iter/s)": 0.054264
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.6285018270401949,
|
| 1351 |
+
"grad_norm": 0.45468801259994507,
|
| 1352 |
+
"learning_rate": 3.323619987579914e-05,
|
| 1353 |
+
"loss": 0.6177504062652588,
|
| 1354 |
+
"memory(GiB)": 387.42,
|
| 1355 |
+
"step": 645,
|
| 1356 |
+
"token_acc": 0.80891932520461,
|
| 1357 |
+
"train_speed(iter/s)": 0.054261
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.633373934226553,
|
| 1361 |
+
"grad_norm": 0.6319808959960938,
|
| 1362 |
+
"learning_rate": 3.247872154681439e-05,
|
| 1363 |
+
"loss": 0.5958673000335694,
|
| 1364 |
+
"memory(GiB)": 387.42,
|
| 1365 |
+
"step": 650,
|
| 1366 |
+
"token_acc": 0.8096597145993414,
|
| 1367 |
+
"train_speed(iter/s)": 0.054221
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.6382460414129111,
|
| 1371 |
+
"grad_norm": 0.4812871217727661,
|
| 1372 |
+
"learning_rate": 3.172580020933442e-05,
|
| 1373 |
+
"loss": 0.5768674850463867,
|
| 1374 |
+
"memory(GiB)": 387.42,
|
| 1375 |
+
"step": 655,
|
| 1376 |
+
"token_acc": 0.8165027102991367,
|
| 1377 |
+
"train_speed(iter/s)": 0.054185
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 0.6431181485992692,
|
| 1381 |
+
"grad_norm": 0.9395345449447632,
|
| 1382 |
+
"learning_rate": 3.097763168560741e-05,
|
| 1383 |
+
"loss": 0.674397611618042,
|
| 1384 |
+
"memory(GiB)": 387.42,
|
| 1385 |
+
"step": 660,
|
| 1386 |
+
"token_acc": 0.7806563039723662,
|
| 1387 |
+
"train_speed(iter/s)": 0.054211
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"epoch": 0.6479902557856273,
|
| 1391 |
+
"grad_norm": 0.5097836852073669,
|
| 1392 |
+
"learning_rate": 3.0234410561754257e-05,
|
| 1393 |
+
"loss": 0.5154216766357422,
|
| 1394 |
+
"memory(GiB)": 387.42,
|
| 1395 |
+
"step": 665,
|
| 1396 |
+
"token_acc": 0.8327868852459016,
|
| 1397 |
+
"train_speed(iter/s)": 0.054197
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 0.6528623629719854,
|
| 1401 |
+
"grad_norm": 0.3545515239238739,
|
| 1402 |
+
"learning_rate": 2.949633013715982e-05,
|
| 1403 |
+
"loss": 0.5994223117828369,
|
| 1404 |
+
"memory(GiB)": 387.42,
|
| 1405 |
+
"step": 670,
|
| 1406 |
+
"token_acc": 0.8076275080410477,
|
| 1407 |
+
"train_speed(iter/s)": 0.054247
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 0.6577344701583435,
|
| 1411 |
+
"grad_norm": 0.9892140030860901,
|
| 1412 |
+
"learning_rate": 2.8763582374199126e-05,
|
| 1413 |
+
"loss": 0.5891304969787597,
|
| 1414 |
+
"memory(GiB)": 387.42,
|
| 1415 |
+
"step": 675,
|
| 1416 |
+
"token_acc": 0.8036573628488932,
|
| 1417 |
+
"train_speed(iter/s)": 0.054243
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 0.6626065773447016,
|
| 1421 |
+
"grad_norm": 0.5605654716491699,
|
| 1422 |
+
"learning_rate": 2.8036357848311012e-05,
|
| 1423 |
+
"loss": 0.5478427410125732,
|
| 1424 |
+
"memory(GiB)": 387.42,
|
| 1425 |
+
"step": 680,
|
| 1426 |
+
"token_acc": 0.8287547623821937,
|
| 1427 |
+
"train_speed(iter/s)": 0.054281
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.6674786845310596,
|
| 1431 |
+
"grad_norm": 0.4100501239299774,
|
| 1432 |
+
"learning_rate": 2.7314845698432805e-05,
|
| 1433 |
+
"loss": 0.6083401203155517,
|
| 1434 |
+
"memory(GiB)": 387.42,
|
| 1435 |
+
"step": 685,
|
| 1436 |
+
"token_acc": 0.7989271180170181,
|
| 1437 |
+
"train_speed(iter/s)": 0.054288
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.6723507917174177,
|
| 1441 |
+
"grad_norm": 0.4639231562614441,
|
| 1442 |
+
"learning_rate": 2.659923357780828e-05,
|
| 1443 |
+
"loss": 0.5717390060424805,
|
| 1444 |
+
"memory(GiB)": 387.42,
|
| 1445 |
+
"step": 690,
|
| 1446 |
+
"token_acc": 0.8201791448369106,
|
| 1447 |
+
"train_speed(iter/s)": 0.054301
|
| 1448 |
+
},
|
| 1449 |
+
{
|
| 1450 |
+
"epoch": 0.6772228989037758,
|
| 1451 |
+
"grad_norm": 0.30558013916015625,
|
| 1452 |
+
"learning_rate": 2.5889707605182347e-05,
|
| 1453 |
+
"loss": 0.4964598178863525,
|
| 1454 |
+
"memory(GiB)": 387.42,
|
| 1455 |
+
"step": 695,
|
| 1456 |
+
"token_acc": 0.8518634024637455,
|
| 1457 |
+
"train_speed(iter/s)": 0.054314
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 0.682095006090134,
|
| 1461 |
+
"grad_norm": 0.490887314081192,
|
| 1462 |
+
"learning_rate": 2.518645231639457e-05,
|
| 1463 |
+
"loss": 0.6779924392700195,
|
| 1464 |
+
"memory(GiB)": 387.42,
|
| 1465 |
+
"step": 700,
|
| 1466 |
+
"token_acc": 0.7798953662182362,
|
| 1467 |
+
"train_speed(iter/s)": 0.054375
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 0.682095006090134,
|
| 1471 |
+
"eval_loss": 0.587890625,
|
| 1472 |
+
"eval_runtime": 6.016,
|
| 1473 |
+
"eval_samples_per_second": 0.665,
|
| 1474 |
+
"eval_steps_per_second": 0.665,
|
| 1475 |
+
"step": 700
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 0.6869671132764921,
|
| 1479 |
+
"grad_norm": 0.9540379047393799,
|
| 1480 |
+
"learning_rate": 2.4489650616384507e-05,
|
| 1481 |
+
"loss": 0.5919107437133789,
|
| 1482 |
+
"memory(GiB)": 387.42,
|
| 1483 |
+
"step": 705,
|
| 1484 |
+
"token_acc": 0.8063427800269906,
|
| 1485 |
+
"train_speed(iter/s)": 0.054286
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 0.6918392204628502,
|
| 1489 |
+
"grad_norm": 0.4385371208190918,
|
| 1490 |
+
"learning_rate": 2.3799483731621237e-05,
|
| 1491 |
+
"loss": 0.5554671287536621,
|
| 1492 |
+
"memory(GiB)": 387.42,
|
| 1493 |
+
"step": 710,
|
| 1494 |
+
"token_acc": 0.8227891742802965,
|
| 1495 |
+
"train_speed(iter/s)": 0.054309
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 0.6967113276492083,
|
| 1499 |
+
"grad_norm": 0.37225764989852905,
|
| 1500 |
+
"learning_rate": 2.311613116296929e-05,
|
| 1501 |
+
"loss": 0.5223379611968995,
|
| 1502 |
+
"memory(GiB)": 387.42,
|
| 1503 |
+
"step": 715,
|
| 1504 |
+
"token_acc": 0.8422697368421053,
|
| 1505 |
+
"train_speed(iter/s)": 0.054303
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 0.7015834348355664,
|
| 1509 |
+
"grad_norm": 0.6227976083755493,
|
| 1510 |
+
"learning_rate": 2.2439770639003627e-05,
|
| 1511 |
+
"loss": 0.5609029769897461,
|
| 1512 |
+
"memory(GiB)": 387.42,
|
| 1513 |
+
"step": 720,
|
| 1514 |
+
"token_acc": 0.8244803695150116,
|
| 1515 |
+
"train_speed(iter/s)": 0.054309
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 0.7064555420219245,
|
| 1519 |
+
"grad_norm": 0.4218509793281555,
|
| 1520 |
+
"learning_rate": 2.177057806978522e-05,
|
| 1521 |
+
"loss": 0.5789398193359375,
|
| 1522 |
+
"memory(GiB)": 387.42,
|
| 1523 |
+
"step": 725,
|
| 1524 |
+
"token_acc": 0.8195275590551181,
|
| 1525 |
+
"train_speed(iter/s)": 0.054317
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 0.7113276492082826,
|
| 1529 |
+
"grad_norm": 0.5081908106803894,
|
| 1530 |
+
"learning_rate": 2.110872750110996e-05,
|
| 1531 |
+
"loss": 0.49318413734436034,
|
| 1532 |
+
"memory(GiB)": 387.42,
|
| 1533 |
+
"step": 730,
|
| 1534 |
+
"token_acc": 0.8306063522617901,
|
| 1535 |
+
"train_speed(iter/s)": 0.05436
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.7161997563946407,
|
| 1539 |
+
"grad_norm": 0.6738778352737427,
|
| 1540 |
+
"learning_rate": 2.045439106924217e-05,
|
| 1541 |
+
"loss": 0.55146803855896,
|
| 1542 |
+
"memory(GiB)": 387.42,
|
| 1543 |
+
"step": 735,
|
| 1544 |
+
"token_acc": 0.8200392927308447,
|
| 1545 |
+
"train_speed(iter/s)": 0.054367
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 0.7210718635809987,
|
| 1549 |
+
"grad_norm": 0.43147921562194824,
|
| 1550 |
+
"learning_rate": 1.980773895614481e-05,
|
| 1551 |
+
"loss": 0.574643898010254,
|
| 1552 |
+
"memory(GiB)": 387.42,
|
| 1553 |
+
"step": 740,
|
| 1554 |
+
"token_acc": 0.8172221384406575,
|
| 1555 |
+
"train_speed(iter/s)": 0.054386
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 0.7259439707673568,
|
| 1559 |
+
"grad_norm": 0.5750350952148438,
|
| 1560 |
+
"learning_rate": 1.9168939345218095e-05,
|
| 1561 |
+
"loss": 0.5682173728942871,
|
| 1562 |
+
"memory(GiB)": 387.42,
|
| 1563 |
+
"step": 745,
|
| 1564 |
+
"token_acc": 0.8214421252371916,
|
| 1565 |
+
"train_speed(iter/s)": 0.054395
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 0.730816077953715,
|
| 1569 |
+
"grad_norm": 0.461907297372818,
|
| 1570 |
+
"learning_rate": 1.8538158377557702e-05,
|
| 1571 |
+
"loss": 0.5272111415863037,
|
| 1572 |
+
"memory(GiB)": 387.42,
|
| 1573 |
+
"step": 750,
|
| 1574 |
+
"token_acc": 0.8257032542746828,
|
| 1575 |
+
"train_speed(iter/s)": 0.054421
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 0.7356881851400731,
|
| 1579 |
+
"grad_norm": 0.794235348701477,
|
| 1580 |
+
"learning_rate": 1.791556010874434e-05,
|
| 1581 |
+
"loss": 0.6292970180511475,
|
| 1582 |
+
"memory(GiB)": 387.42,
|
| 1583 |
+
"step": 755,
|
| 1584 |
+
"token_acc": 0.810012836970475,
|
| 1585 |
+
"train_speed(iter/s)": 0.054353
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 0.7405602923264312,
|
| 1589 |
+
"grad_norm": 0.6189777851104736,
|
| 1590 |
+
"learning_rate": 1.7301306466175533e-05,
|
| 1591 |
+
"loss": 0.5557656288146973,
|
| 1592 |
+
"memory(GiB)": 387.42,
|
| 1593 |
+
"step": 760,
|
| 1594 |
+
"token_acc": 0.8259242957746479,
|
| 1595 |
+
"train_speed(iter/s)": 0.054349
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 0.7454323995127893,
|
| 1599 |
+
"grad_norm": 0.4845249056816101,
|
| 1600 |
+
"learning_rate": 1.6695557206951144e-05,
|
| 1601 |
+
"loss": 0.49696760177612304,
|
| 1602 |
+
"memory(GiB)": 389.68,
|
| 1603 |
+
"step": 765,
|
| 1604 |
+
"token_acc": 0.8422638261243813,
|
| 1605 |
+
"train_speed(iter/s)": 0.054323
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.7503045066991474,
|
| 1609 |
+
"grad_norm": 0.4710843563079834,
|
| 1610 |
+
"learning_rate": 1.6098469876323093e-05,
|
| 1611 |
+
"loss": 0.47034273147583006,
|
| 1612 |
+
"memory(GiB)": 389.68,
|
| 1613 |
+
"step": 770,
|
| 1614 |
+
"token_acc": 0.8487571701720842,
|
| 1615 |
+
"train_speed(iter/s)": 0.05434
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 0.7551766138855055,
|
| 1619 |
+
"grad_norm": 0.45380252599716187,
|
| 1620 |
+
"learning_rate": 1.551019976672058e-05,
|
| 1621 |
+
"loss": 0.5777853488922119,
|
| 1622 |
+
"memory(GiB)": 389.68,
|
| 1623 |
+
"step": 775,
|
| 1624 |
+
"token_acc": 0.8110020910406949,
|
| 1625 |
+
"train_speed(iter/s)": 0.054377
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 0.7600487210718636,
|
| 1629 |
+
"grad_norm": 0.5304797291755676,
|
| 1630 |
+
"learning_rate": 1.4930899877361015e-05,
|
| 1631 |
+
"loss": 0.5180749416351318,
|
| 1632 |
+
"memory(GiB)": 389.68,
|
| 1633 |
+
"step": 780,
|
| 1634 |
+
"token_acc": 0.8334659769200159,
|
| 1635 |
+
"train_speed(iter/s)": 0.05443
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 0.7649208282582217,
|
| 1639 |
+
"grad_norm": 0.447553426027298,
|
| 1640 |
+
"learning_rate": 1.4360720874457607e-05,
|
| 1641 |
+
"loss": 0.5336573123931885,
|
| 1642 |
+
"memory(GiB)": 389.68,
|
| 1643 |
+
"step": 785,
|
| 1644 |
+
"token_acc": 0.8346641615782058,
|
| 1645 |
+
"train_speed(iter/s)": 0.054438
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 0.7697929354445798,
|
| 1649 |
+
"grad_norm": 0.5468970537185669,
|
| 1650 |
+
"learning_rate": 1.3799811052033467e-05,
|
| 1651 |
+
"loss": 0.6092133522033691,
|
| 1652 |
+
"memory(GiB)": 389.68,
|
| 1653 |
+
"step": 790,
|
| 1654 |
+
"token_acc": 0.7997620261771206,
|
| 1655 |
+
"train_speed(iter/s)": 0.054456
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 0.7746650426309378,
|
| 1659 |
+
"grad_norm": 0.6424246430397034,
|
| 1660 |
+
"learning_rate": 1.3248316293352946e-05,
|
| 1661 |
+
"loss": 0.6084504127502441,
|
| 1662 |
+
"memory(GiB)": 389.68,
|
| 1663 |
+
"step": 795,
|
| 1664 |
+
"token_acc": 0.8091853471842537,
|
| 1665 |
+
"train_speed(iter/s)": 0.05451
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 0.7795371498172959,
|
| 1669 |
+
"grad_norm": 0.5339289903640747,
|
| 1670 |
+
"learning_rate": 1.2706380032979691e-05,
|
| 1671 |
+
"loss": 0.535353136062622,
|
| 1672 |
+
"memory(GiB)": 389.68,
|
| 1673 |
+
"step": 800,
|
| 1674 |
+
"token_acc": 0.8231229847996315,
|
| 1675 |
+
"train_speed(iter/s)": 0.054509
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.7795371498172959,
|
| 1679 |
+
"eval_loss": 0.587626039981842,
|
| 1680 |
+
"eval_runtime": 6.1485,
|
| 1681 |
+
"eval_samples_per_second": 0.651,
|
| 1682 |
+
"eval_steps_per_second": 0.651,
|
| 1683 |
+
"step": 800
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.784409257003654,
|
| 1687 |
+
"grad_norm": 0.47259068489074707,
|
| 1688 |
+
"learning_rate": 1.2174143219471878e-05,
|
| 1689 |
+
"loss": 0.6263217449188232,
|
| 1690 |
+
"memory(GiB)": 389.68,
|
| 1691 |
+
"step": 805,
|
| 1692 |
+
"token_acc": 0.7991557070953077,
|
| 1693 |
+
"train_speed(iter/s)": 0.054434
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 0.7892813641900122,
|
| 1697 |
+
"grad_norm": 0.5547453761100769,
|
| 1698 |
+
"learning_rate": 1.1651744278723687e-05,
|
| 1699 |
+
"loss": 0.5090929985046386,
|
| 1700 |
+
"memory(GiB)": 389.68,
|
| 1701 |
+
"step": 810,
|
| 1702 |
+
"token_acc": 0.8354404976921533,
|
| 1703 |
+
"train_speed(iter/s)": 0.054448
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.7941534713763703,
|
| 1707 |
+
"grad_norm": 0.4848991930484772,
|
| 1708 |
+
"learning_rate": 1.1139319077963178e-05,
|
| 1709 |
+
"loss": 0.5273432254791259,
|
| 1710 |
+
"memory(GiB)": 389.68,
|
| 1711 |
+
"step": 815,
|
| 1712 |
+
"token_acc": 0.8295368261199696,
|
| 1713 |
+
"train_speed(iter/s)": 0.054475
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 0.7990255785627284,
|
| 1717 |
+
"grad_norm": 0.5590830445289612,
|
| 1718 |
+
"learning_rate": 1.0637000890415388e-05,
|
| 1719 |
+
"loss": 0.6279808044433594,
|
| 1720 |
+
"memory(GiB)": 389.68,
|
| 1721 |
+
"step": 820,
|
| 1722 |
+
"token_acc": 0.8061934585942937,
|
| 1723 |
+
"train_speed(iter/s)": 0.054494
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"epoch": 0.8038976857490865,
|
| 1727 |
+
"grad_norm": 1.119874358177185,
|
| 1728 |
+
"learning_rate": 1.0144920360640303e-05,
|
| 1729 |
+
"loss": 0.6255881309509277,
|
| 1730 |
+
"memory(GiB)": 389.68,
|
| 1731 |
+
"step": 825,
|
| 1732 |
+
"token_acc": 0.8063498323802012,
|
| 1733 |
+
"train_speed(iter/s)": 0.0545
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 0.8087697929354446,
|
| 1737 |
+
"grad_norm": 0.4502837359905243,
|
| 1738 |
+
"learning_rate": 9.663205470554276e-06,
|
| 1739 |
+
"loss": 0.5530724048614502,
|
| 1740 |
+
"memory(GiB)": 389.68,
|
| 1741 |
+
"step": 830,
|
| 1742 |
+
"token_acc": 0.8286991062562066,
|
| 1743 |
+
"train_speed(iter/s)": 0.054498
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 0.8136419001218027,
|
| 1747 |
+
"grad_norm": 0.47327640652656555,
|
| 1748 |
+
"learning_rate": 9.19198150614417e-06,
|
| 1749 |
+
"loss": 0.6426435470581054,
|
| 1750 |
+
"memory(GiB)": 389.68,
|
| 1751 |
+
"step": 835,
|
| 1752 |
+
"token_acc": 0.7995495495495496,
|
| 1753 |
+
"train_speed(iter/s)": 0.054482
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.8185140073081608,
|
| 1757 |
+
"grad_norm": 0.45425912737846375,
|
| 1758 |
+
"learning_rate": 8.73137102488249e-06,
|
| 1759 |
+
"loss": 0.5113016128540039,
|
| 1760 |
+
"memory(GiB)": 389.68,
|
| 1761 |
+
"step": 840,
|
| 1762 |
+
"token_acc": 0.8368200836820083,
|
| 1763 |
+
"train_speed(iter/s)": 0.054528
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 0.8233861144945189,
|
| 1767 |
+
"grad_norm": 0.5594798922538757,
|
| 1768 |
+
"learning_rate": 8.28149382385231e-06,
|
| 1769 |
+
"loss": 0.5977861881256104,
|
| 1770 |
+
"memory(GiB)": 389.68,
|
| 1771 |
+
"step": 845,
|
| 1772 |
+
"token_acc": 0.8159670164917541,
|
| 1773 |
+
"train_speed(iter/s)": 0.054545
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.8282582216808769,
|
| 1777 |
+
"grad_norm": 0.38594865798950195,
|
| 1778 |
+
"learning_rate": 7.842466908590006e-06,
|
| 1779 |
+
"loss": 0.5546538829803467,
|
| 1780 |
+
"memory(GiB)": 389.68,
|
| 1781 |
+
"step": 850,
|
| 1782 |
+
"token_acc": 0.8362763915547025,
|
| 1783 |
+
"train_speed(iter/s)": 0.05454
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 0.833130328867235,
|
| 1787 |
+
"grad_norm": 0.6128694415092468,
|
| 1788 |
+
"learning_rate": 7.414404462654051e-06,
|
| 1789 |
+
"loss": 0.5578857898712158,
|
| 1790 |
+
"memory(GiB)": 389.68,
|
| 1791 |
+
"step": 855,
|
| 1792 |
+
"token_acc": 0.8173973075595443,
|
| 1793 |
+
"train_speed(iter/s)": 0.054466
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 0.8380024360535931,
|
| 1797 |
+
"grad_norm": 0.5973862409591675,
|
| 1798 |
+
"learning_rate": 6.997417817927865e-06,
|
| 1799 |
+
"loss": 0.6116644382476807,
|
| 1800 |
+
"memory(GiB)": 389.68,
|
| 1801 |
+
"step": 860,
|
| 1802 |
+
"token_acc": 0.8100558659217877,
|
| 1803 |
+
"train_speed(iter/s)": 0.054467
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 0.8428745432399513,
|
| 1807 |
+
"grad_norm": 0.5695779323577881,
|
| 1808 |
+
"learning_rate": 6.591615425664144e-06,
|
| 1809 |
+
"loss": 0.6063879013061524,
|
| 1810 |
+
"memory(GiB)": 389.68,
|
| 1811 |
+
"step": 865,
|
| 1812 |
+
"token_acc": 0.8113871180479226,
|
| 1813 |
+
"train_speed(iter/s)": 0.054502
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 0.8477466504263094,
|
| 1817 |
+
"grad_norm": 0.37414440512657166,
|
| 1818 |
+
"learning_rate": 6.197102828278611e-06,
|
| 1819 |
+
"loss": 0.5134734153747559,
|
| 1820 |
+
"memory(GiB)": 389.68,
|
| 1821 |
+
"step": 870,
|
| 1822 |
+
"token_acc": 0.8304152076038019,
|
| 1823 |
+
"train_speed(iter/s)": 0.054524
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 0.8526187576126675,
|
| 1827 |
+
"grad_norm": 0.8222331404685974,
|
| 1828 |
+
"learning_rate": 5.813982631900122e-06,
|
| 1829 |
+
"loss": 0.5653984069824218,
|
| 1830 |
+
"memory(GiB)": 389.68,
|
| 1831 |
+
"step": 875,
|
| 1832 |
+
"token_acc": 0.8229976496112819,
|
| 1833 |
+
"train_speed(iter/s)": 0.054534
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 0.8574908647990256,
|
| 1837 |
+
"grad_norm": 0.3609310984611511,
|
| 1838 |
+
"learning_rate": 5.442354479684558e-06,
|
| 1839 |
+
"loss": 0.49175424575805665,
|
| 1840 |
+
"memory(GiB)": 389.68,
|
| 1841 |
+
"step": 880,
|
| 1842 |
+
"token_acc": 0.8409646976581615,
|
| 1843 |
+
"train_speed(iter/s)": 0.054533
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.8623629719853837,
|
| 1847 |
+
"grad_norm": 0.6293960213661194,
|
| 1848 |
+
"learning_rate": 5.082315025899315e-06,
|
| 1849 |
+
"loss": 0.604953384399414,
|
| 1850 |
+
"memory(GiB)": 389.68,
|
| 1851 |
+
"step": 885,
|
| 1852 |
+
"token_acc": 0.8073544433094995,
|
| 1853 |
+
"train_speed(iter/s)": 0.05455
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"epoch": 0.8672350791717418,
|
| 1857 |
+
"grad_norm": 0.4242098331451416,
|
| 1858 |
+
"learning_rate": 4.733957910785114e-06,
|
| 1859 |
+
"loss": 0.4986411571502686,
|
| 1860 |
+
"memory(GiB)": 389.68,
|
| 1861 |
+
"step": 890,
|
| 1862 |
+
"token_acc": 0.8444040036396724,
|
| 1863 |
+
"train_speed(iter/s)": 0.054562
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 0.8721071863580999,
|
| 1867 |
+
"grad_norm": 0.5025205612182617,
|
| 1868 |
+
"learning_rate": 4.397373736201782e-06,
|
| 1869 |
+
"loss": 0.5355000495910645,
|
| 1870 |
+
"memory(GiB)": 389.68,
|
| 1871 |
+
"step": 895,
|
| 1872 |
+
"token_acc": 0.8340460526315789,
|
| 1873 |
+
"train_speed(iter/s)": 0.054564
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 0.876979293544458,
|
| 1877 |
+
"grad_norm": 0.42587506771087646,
|
| 1878 |
+
"learning_rate": 4.072650042064174e-06,
|
| 1879 |
+
"loss": 0.6113440513610839,
|
| 1880 |
+
"memory(GiB)": 389.68,
|
| 1881 |
+
"step": 900,
|
| 1882 |
+
"token_acc": 0.8042306924765515,
|
| 1883 |
+
"train_speed(iter/s)": 0.054571
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 0.876979293544458,
|
| 1887 |
+
"eval_loss": 0.5867875814437866,
|
| 1888 |
+
"eval_runtime": 6.1618,
|
| 1889 |
+
"eval_samples_per_second": 0.649,
|
| 1890 |
+
"eval_steps_per_second": 0.649,
|
| 1891 |
+
"step": 900
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 0.881851400730816,
|
| 1895 |
+
"grad_norm": 0.6062163710594177,
|
| 1896 |
+
"learning_rate": 3.759871283574562e-06,
|
| 1897 |
+
"loss": 0.5853659629821777,
|
| 1898 |
+
"memory(GiB)": 389.68,
|
| 1899 |
+
"step": 905,
|
| 1900 |
+
"token_acc": 0.8163235076284995,
|
| 1901 |
+
"train_speed(iter/s)": 0.054495
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 0.8867235079171741,
|
| 1905 |
+
"grad_norm": 0.5810290575027466,
|
| 1906 |
+
"learning_rate": 3.4591188092571893e-06,
|
| 1907 |
+
"loss": 0.5189132213592529,
|
| 1908 |
+
"memory(GiB)": 389.68,
|
| 1909 |
+
"step": 910,
|
| 1910 |
+
"token_acc": 0.848421052631579,
|
| 1911 |
+
"train_speed(iter/s)": 0.054517
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 0.8915956151035322,
|
| 1915 |
+
"grad_norm": 0.5703849196434021,
|
| 1916 |
+
"learning_rate": 3.1704708398009486e-06,
|
| 1917 |
+
"loss": 0.5976828575134278,
|
| 1918 |
+
"memory(GiB)": 389.68,
|
| 1919 |
+
"step": 915,
|
| 1920 |
+
"token_acc": 0.808837066584842,
|
| 1921 |
+
"train_speed(iter/s)": 0.05451
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 0.8964677222898904,
|
| 1925 |
+
"grad_norm": 0.5777165293693542,
|
| 1926 |
+
"learning_rate": 2.894002447715399e-06,
|
| 1927 |
+
"loss": 0.5165195465087891,
|
| 1928 |
+
"memory(GiB)": 389.68,
|
| 1929 |
+
"step": 920,
|
| 1930 |
+
"token_acc": 0.8424015009380863,
|
| 1931 |
+
"train_speed(iter/s)": 0.054567
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 0.9013398294762485,
|
| 1935 |
+
"grad_norm": 0.48375067114830017,
|
| 1936 |
+
"learning_rate": 2.6297855378057623e-06,
|
| 1937 |
+
"loss": 0.46347522735595703,
|
| 1938 |
+
"memory(GiB)": 389.68,
|
| 1939 |
+
"step": 925,
|
| 1940 |
+
"token_acc": 0.8408729585200173,
|
| 1941 |
+
"train_speed(iter/s)": 0.054561
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.9062119366626066,
|
| 1945 |
+
"grad_norm": 0.4930781126022339,
|
| 1946 |
+
"learning_rate": 2.3778888284716193e-06,
|
| 1947 |
+
"loss": 0.6031323909759522,
|
| 1948 |
+
"memory(GiB)": 389.68,
|
| 1949 |
+
"step": 930,
|
| 1950 |
+
"token_acc": 0.8058429701765064,
|
| 1951 |
+
"train_speed(iter/s)": 0.054553
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"epoch": 0.9110840438489647,
|
| 1955 |
+
"grad_norm": 0.42932575941085815,
|
| 1956 |
+
"learning_rate": 2.138377833834404e-06,
|
| 1957 |
+
"loss": 0.5199082851409912,
|
| 1958 |
+
"memory(GiB)": 389.68,
|
| 1959 |
+
"step": 935,
|
| 1960 |
+
"token_acc": 0.837616269903831,
|
| 1961 |
+
"train_speed(iter/s)": 0.054552
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 0.9159561510353228,
|
| 1965 |
+
"grad_norm": 0.6615188717842102,
|
| 1966 |
+
"learning_rate": 1.9113148466983254e-06,
|
| 1967 |
+
"loss": 0.6138844013214111,
|
| 1968 |
+
"memory(GiB)": 389.68,
|
| 1969 |
+
"step": 940,
|
| 1970 |
+
"token_acc": 0.8027118644067797,
|
| 1971 |
+
"train_speed(iter/s)": 0.054582
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 0.9208282582216809,
|
| 1975 |
+
"grad_norm": 0.41028302907943726,
|
| 1976 |
+
"learning_rate": 1.696758922348979e-06,
|
| 1977 |
+
"loss": 0.5526364803314209,
|
| 1978 |
+
"memory(GiB)": 389.68,
|
| 1979 |
+
"step": 945,
|
| 1980 |
+
"token_acc": 0.8190247252747253,
|
| 1981 |
+
"train_speed(iter/s)": 0.054578
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 0.925700365408039,
|
| 1985 |
+
"grad_norm": 0.48014047741889954,
|
| 1986 |
+
"learning_rate": 1.4947658631941309e-06,
|
| 1987 |
+
"loss": 0.49515771865844727,
|
| 1988 |
+
"memory(GiB)": 389.68,
|
| 1989 |
+
"step": 950,
|
| 1990 |
+
"token_acc": 0.832800851970181,
|
| 1991 |
+
"train_speed(iter/s)": 0.054557
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 0.9305724725943971,
|
| 1995 |
+
"grad_norm": 0.6173512935638428,
|
| 1996 |
+
"learning_rate": 1.3053882042503796e-06,
|
| 1997 |
+
"loss": 0.5243947505950928,
|
| 1998 |
+
"memory(GiB)": 389.68,
|
| 1999 |
+
"step": 955,
|
| 2000 |
+
"token_acc": 0.8282737560625112,
|
| 2001 |
+
"train_speed(iter/s)": 0.054472
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 0.9354445797807551,
|
| 2005 |
+
"grad_norm": 0.6899262070655823,
|
| 2006 |
+
"learning_rate": 1.1286751994797284e-06,
|
| 2007 |
+
"loss": 0.636317253112793,
|
| 2008 |
+
"memory(GiB)": 389.68,
|
| 2009 |
+
"step": 960,
|
| 2010 |
+
"token_acc": 0.8041509433962264,
|
| 2011 |
+
"train_speed(iter/s)": 0.05449
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.9403166869671132,
|
| 2015 |
+
"grad_norm": 0.538864016532898,
|
| 2016 |
+
"learning_rate": 9.646728089794167e-07,
|
| 2017 |
+
"loss": 0.5281119823455811,
|
| 2018 |
+
"memory(GiB)": 389.68,
|
| 2019 |
+
"step": 965,
|
| 2020 |
+
"token_acc": 0.828132906054984,
|
| 2021 |
+
"train_speed(iter/s)": 0.054472
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 0.9451887941534713,
|
| 2025 |
+
"grad_norm": 0.7353665828704834,
|
| 2026 |
+
"learning_rate": 8.134236870284861e-07,
|
| 2027 |
+
"loss": 0.6087577819824219,
|
| 2028 |
+
"memory(GiB)": 389.68,
|
| 2029 |
+
"step": 970,
|
| 2030 |
+
"token_acc": 0.8098674274207082,
|
| 2031 |
+
"train_speed(iter/s)": 0.054485
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 0.9500609013398295,
|
| 2035 |
+
"grad_norm": 0.7473301887512207,
|
| 2036 |
+
"learning_rate": 6.749671709941008e-07,
|
| 2037 |
+
"loss": 0.6141918182373047,
|
| 2038 |
+
"memory(GiB)": 389.68,
|
| 2039 |
+
"step": 975,
|
| 2040 |
+
"token_acc": 0.8016149752248118,
|
| 2041 |
+
"train_speed(iter/s)": 0.054518
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 0.9549330085261876,
|
| 2045 |
+
"grad_norm": 0.6487853527069092,
|
| 2046 |
+
"learning_rate": 5.493392711005796e-07,
|
| 2047 |
+
"loss": 0.5959615707397461,
|
| 2048 |
+
"memory(GiB)": 389.68,
|
| 2049 |
+
"step": 980,
|
| 2050 |
+
"token_acc": 0.8156642881413524,
|
| 2051 |
+
"train_speed(iter/s)": 0.054561
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 0.9598051157125457,
|
| 2055 |
+
"grad_norm": 0.678453803062439,
|
| 2056 |
+
"learning_rate": 4.365726610637222e-07,
|
| 2057 |
+
"loss": 0.5411821842193604,
|
| 2058 |
+
"memory(GiB)": 389.68,
|
| 2059 |
+
"step": 985,
|
| 2060 |
+
"token_acc": 0.8313556274721323,
|
| 2061 |
+
"train_speed(iter/s)": 0.054544
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 0.9646772228989038,
|
| 2065 |
+
"grad_norm": 0.5119591355323792,
|
| 2066 |
+
"learning_rate": 3.366966695929119e-07,
|
| 2067 |
+
"loss": 0.49676513671875,
|
| 2068 |
+
"memory(GiB)": 389.68,
|
| 2069 |
+
"step": 990,
|
| 2070 |
+
"token_acc": 0.8351805505899178,
|
| 2071 |
+
"train_speed(iter/s)": 0.054553
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"epoch": 0.9695493300852619,
|
| 2075 |
+
"grad_norm": 0.6289726495742798,
|
| 2076 |
+
"learning_rate": 2.4973727276323965e-07,
|
| 2077 |
+
"loss": 0.60072922706604,
|
| 2078 |
+
"memory(GiB)": 389.68,
|
| 2079 |
+
"step": 995,
|
| 2080 |
+
"token_acc": 0.8124610591900312,
|
| 2081 |
+
"train_speed(iter/s)": 0.054575
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.97442143727162,
|
| 2085 |
+
"grad_norm": 0.5490319132804871,
|
| 2086 |
+
"learning_rate": 1.7571708725953596e-07,
|
| 2087 |
+
"loss": 0.5364939212799072,
|
| 2088 |
+
"memory(GiB)": 389.68,
|
| 2089 |
+
"step": 1000,
|
| 2090 |
+
"token_acc": 0.8235892221657346,
|
| 2091 |
+
"train_speed(iter/s)": 0.054556
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 0.97442143727162,
|
| 2095 |
+
"eval_loss": 0.5838146805763245,
|
| 2096 |
+
"eval_runtime": 6.1207,
|
| 2097 |
+
"eval_samples_per_second": 0.654,
|
| 2098 |
+
"eval_steps_per_second": 0.654,
|
| 2099 |
+
"step": 1000
|
| 2100 |
+
},
|
| 2101 |
+
{
|
| 2102 |
+
"epoch": 0.9792935444579781,
|
| 2103 |
+
"grad_norm": 0.4941563010215759,
|
| 2104 |
+
"learning_rate": 1.1465536449415393e-07,
|
| 2105 |
+
"loss": 0.5735920906066895,
|
| 2106 |
+
"memory(GiB)": 389.68,
|
| 2107 |
+
"step": 1005,
|
| 2108 |
+
"token_acc": 0.8156670746634027,
|
| 2109 |
+
"train_speed(iter/s)": 0.054474
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.9841656516443362,
|
| 2113 |
+
"grad_norm": 0.5679388046264648,
|
| 2114 |
+
"learning_rate": 6.656798560001343e-08,
|
| 2115 |
+
"loss": 0.5337845325469971,
|
| 2116 |
+
"memory(GiB)": 389.68,
|
| 2117 |
+
"step": 1010,
|
| 2118 |
+
"token_acc": 0.8183527641970666,
|
| 2119 |
+
"train_speed(iter/s)": 0.054492
|
| 2120 |
+
},
|
| 2121 |
+
{
|
| 2122 |
+
"epoch": 0.9890377588306942,
|
| 2123 |
+
"grad_norm": 0.43481603264808655,
|
| 2124 |
+
"learning_rate": 3.146745730015499e-08,
|
| 2125 |
+
"loss": 0.5338433265686036,
|
| 2126 |
+
"memory(GiB)": 389.68,
|
| 2127 |
+
"step": 1015,
|
| 2128 |
+
"token_acc": 0.8283907544701264,
|
| 2129 |
+
"train_speed(iter/s)": 0.054525
|
| 2130 |
+
},
|
| 2131 |
+
{
|
| 2132 |
+
"epoch": 0.9939098660170523,
|
| 2133 |
+
"grad_norm": 0.44339102506637573,
|
| 2134 |
+
"learning_rate": 9.362908654986235e-09,
|
| 2135 |
+
"loss": 0.5187356472015381,
|
| 2136 |
+
"memory(GiB)": 389.68,
|
| 2137 |
+
"step": 1020,
|
| 2138 |
+
"token_acc": 0.8316270566727605,
|
| 2139 |
+
"train_speed(iter/s)": 0.054538
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 0.9987819732034104,
|
| 2143 |
+
"grad_norm": 0.7172895669937134,
|
| 2144 |
+
"learning_rate": 2.6008868793114817e-10,
|
| 2145 |
+
"loss": 0.5243105888366699,
|
| 2146 |
+
"memory(GiB)": 389.68,
|
| 2147 |
+
"step": 1025,
|
| 2148 |
+
"token_acc": 0.8462152666879591,
|
| 2149 |
+
"train_speed(iter/s)": 0.054564
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 0.9997563946406821,
|
| 2153 |
+
"eval_loss": 0.5837547183036804,
|
| 2154 |
+
"eval_runtime": 6.0694,
|
| 2155 |
+
"eval_samples_per_second": 0.659,
|
| 2156 |
+
"eval_steps_per_second": 0.659,
|
| 2157 |
+
"step": 1026
|
| 2158 |
+
}
|
| 2159 |
+
],
|
| 2160 |
+
"logging_steps": 5,
|
| 2161 |
+
"max_steps": 1026,
|
| 2162 |
+
"num_input_tokens_seen": 0,
|
| 2163 |
+
"num_train_epochs": 1,
|
| 2164 |
+
"save_steps": 50,
|
| 2165 |
+
"stateful_callbacks": {
|
| 2166 |
+
"TrainerControl": {
|
| 2167 |
+
"args": {
|
| 2168 |
+
"should_epoch_stop": false,
|
| 2169 |
+
"should_evaluate": false,
|
| 2170 |
+
"should_log": false,
|
| 2171 |
+
"should_save": true,
|
| 2172 |
+
"should_training_stop": true
|
| 2173 |
+
},
|
| 2174 |
+
"attributes": {}
|
| 2175 |
+
}
|
| 2176 |
+
},
|
| 2177 |
+
"total_flos": 3.095035636732416e+18,
|
| 2178 |
+
"train_batch_size": 1,
|
| 2179 |
+
"trial_name": null,
|
| 2180 |
+
"trial_params": null
|
| 2181 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:150b5a257bdebe994b2e6dfa0c759ef57d68e9a51667b3ba02bf53aa8b0f4ea9
|
| 3 |
+
size 5816
|