140 epoch 555 ds
Browse files- README.md +21 -0
- adapter_config.json +28 -0
- adapter_model.bin +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +971 -0
- training_args.bin +3 -0
README.md
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
| 4 |
+
## Training procedure
|
| 5 |
+
|
| 6 |
+
|
| 7 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 8 |
+
- quant_method: bitsandbytes
|
| 9 |
+
- load_in_8bit: True
|
| 10 |
+
- load_in_4bit: False
|
| 11 |
+
- llm_int8_threshold: 6.0
|
| 12 |
+
- llm_int8_skip_modules: None
|
| 13 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 14 |
+
- llm_int8_has_fp16_weight: False
|
| 15 |
+
- bnb_4bit_quant_type: fp4
|
| 16 |
+
- bnb_4bit_use_double_quant: False
|
| 17 |
+
- bnb_4bit_compute_dtype: float32
|
| 18 |
+
### Framework versions
|
| 19 |
+
|
| 20 |
+
|
| 21 |
+
- PEFT 0.6.0.dev0
|
adapter_config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "codellama/CodeLlama-7b-hf",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"fan_in_fan_out": null,
|
| 7 |
+
"inference_mode": true,
|
| 8 |
+
"init_lora_weights": true,
|
| 9 |
+
"layers_pattern": null,
|
| 10 |
+
"layers_to_transform": null,
|
| 11 |
+
"lora_alpha": 16,
|
| 12 |
+
"lora_dropout": 0.05,
|
| 13 |
+
"modules_to_save": null,
|
| 14 |
+
"peft_type": "LORA",
|
| 15 |
+
"r": 32,
|
| 16 |
+
"rank_pattern": {},
|
| 17 |
+
"revision": null,
|
| 18 |
+
"target_modules": [
|
| 19 |
+
"up_proj",
|
| 20 |
+
"o_proj",
|
| 21 |
+
"q_proj",
|
| 22 |
+
"k_proj",
|
| 23 |
+
"v_proj",
|
| 24 |
+
"down_proj",
|
| 25 |
+
"gate_proj"
|
| 26 |
+
],
|
| 27 |
+
"task_type": "CAUSAL_LM"
|
| 28 |
+
}
|
adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b4bd1b0848329afc323724cb5f5c5c2e8242b4530d3c2fbf9f731c37ecec5f5e
|
| 3 |
+
size 319977229
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5b88eae799662f074f9ac9b776c08827f05d50a1563db15bc68cb9cdd9cbba67
|
| 3 |
+
size 160735647
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b1a9eb3da843c7fb5628d7dfeface97e0eeb5a7219b46b61289345f983637f40
|
| 3 |
+
size 14575
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0900a6fba1d6009e7b51215cdc66f0d705cb3e1d22579b7333dd467a9a360671
|
| 3 |
+
size 627
|
trainer_state.json
ADDED
|
@@ -0,0 +1,971 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8675475120544434,
|
| 3 |
+
"best_model_checkpoint": "./lora-out/checkpoint-60",
|
| 4 |
+
"epoch": 93.33333333333333,
|
| 5 |
+
"eval_steps": 10,
|
| 6 |
+
"global_step": 140,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 1.0,
|
| 13 |
+
"learning_rate": 2e-05,
|
| 14 |
+
"loss": 1.9263,
|
| 15 |
+
"step": 1
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 1.33,
|
| 19 |
+
"learning_rate": 4e-05,
|
| 20 |
+
"loss": 0.6362,
|
| 21 |
+
"step": 2
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 2.0,
|
| 25 |
+
"learning_rate": 6e-05,
|
| 26 |
+
"loss": 1.2827,
|
| 27 |
+
"step": 3
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 2.67,
|
| 31 |
+
"learning_rate": 8e-05,
|
| 32 |
+
"loss": 1.271,
|
| 33 |
+
"step": 4
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 3.0,
|
| 37 |
+
"learning_rate": 0.0001,
|
| 38 |
+
"loss": 0.634,
|
| 39 |
+
"step": 5
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 4.0,
|
| 43 |
+
"learning_rate": 0.00012,
|
| 44 |
+
"loss": 1.8734,
|
| 45 |
+
"step": 6
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 5.0,
|
| 49 |
+
"learning_rate": 0.00014,
|
| 50 |
+
"loss": 1.815,
|
| 51 |
+
"step": 7
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 5.33,
|
| 55 |
+
"learning_rate": 0.00016,
|
| 56 |
+
"loss": 0.5571,
|
| 57 |
+
"step": 8
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 6.0,
|
| 61 |
+
"learning_rate": 0.00018,
|
| 62 |
+
"loss": 1.0428,
|
| 63 |
+
"step": 9
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 6.67,
|
| 67 |
+
"learning_rate": 0.0002,
|
| 68 |
+
"loss": 0.953,
|
| 69 |
+
"step": 10
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 6.67,
|
| 73 |
+
"eval_loss": 1.6378157138824463,
|
| 74 |
+
"eval_runtime": 0.8713,
|
| 75 |
+
"eval_samples_per_second": 6.886,
|
| 76 |
+
"eval_steps_per_second": 3.443,
|
| 77 |
+
"step": 10
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 7.0,
|
| 81 |
+
"learning_rate": 0.00019997482349425066,
|
| 82 |
+
"loss": 0.4287,
|
| 83 |
+
"step": 11
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 8.0,
|
| 87 |
+
"learning_rate": 0.00019989930665413147,
|
| 88 |
+
"loss": 1.2133,
|
| 89 |
+
"step": 12
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 9.0,
|
| 93 |
+
"learning_rate": 0.0001997734875046456,
|
| 94 |
+
"loss": 1.1351,
|
| 95 |
+
"step": 13
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 9.33,
|
| 99 |
+
"learning_rate": 0.00019959742939952392,
|
| 100 |
+
"loss": 0.355,
|
| 101 |
+
"step": 14
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 10.0,
|
| 105 |
+
"learning_rate": 0.00019937122098932428,
|
| 106 |
+
"loss": 0.68,
|
| 107 |
+
"step": 15
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 10.67,
|
| 111 |
+
"learning_rate": 0.00019909497617679348,
|
| 112 |
+
"loss": 0.6661,
|
| 113 |
+
"step": 16
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 11.0,
|
| 117 |
+
"learning_rate": 0.00019876883405951377,
|
| 118 |
+
"loss": 0.3162,
|
| 119 |
+
"step": 17
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 12.0,
|
| 123 |
+
"learning_rate": 0.00019839295885986296,
|
| 124 |
+
"loss": 0.9141,
|
| 125 |
+
"step": 18
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 13.0,
|
| 129 |
+
"learning_rate": 0.00019796753984232358,
|
| 130 |
+
"loss": 0.8809,
|
| 131 |
+
"step": 19
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 13.33,
|
| 135 |
+
"learning_rate": 0.00019749279121818235,
|
| 136 |
+
"loss": 0.2831,
|
| 137 |
+
"step": 20
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 13.33,
|
| 141 |
+
"eval_loss": 1.0763869285583496,
|
| 142 |
+
"eval_runtime": 0.8722,
|
| 143 |
+
"eval_samples_per_second": 6.879,
|
| 144 |
+
"eval_steps_per_second": 3.44,
|
| 145 |
+
"step": 20
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 14.0,
|
| 149 |
+
"learning_rate": 0.0001969689520376687,
|
| 150 |
+
"loss": 0.5415,
|
| 151 |
+
"step": 21
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 14.67,
|
| 155 |
+
"learning_rate": 0.00019639628606958533,
|
| 156 |
+
"loss": 0.5414,
|
| 157 |
+
"step": 22
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 15.0,
|
| 161 |
+
"learning_rate": 0.00019577508166849304,
|
| 162 |
+
"loss": 0.2592,
|
| 163 |
+
"step": 23
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 16.0,
|
| 167 |
+
"learning_rate": 0.00019510565162951537,
|
| 168 |
+
"loss": 0.7754,
|
| 169 |
+
"step": 24
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 17.0,
|
| 173 |
+
"learning_rate": 0.00019438833303083678,
|
| 174 |
+
"loss": 0.7689,
|
| 175 |
+
"step": 25
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 17.33,
|
| 179 |
+
"learning_rate": 0.00019362348706397373,
|
| 180 |
+
"loss": 0.2523,
|
| 181 |
+
"step": 26
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 18.0,
|
| 185 |
+
"learning_rate": 0.0001928114988519039,
|
| 186 |
+
"loss": 0.5015,
|
| 187 |
+
"step": 27
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 18.67,
|
| 191 |
+
"learning_rate": 0.0001919527772551451,
|
| 192 |
+
"loss": 0.5023,
|
| 193 |
+
"step": 28
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 19.0,
|
| 197 |
+
"learning_rate": 0.00019104775466588161,
|
| 198 |
+
"loss": 0.2493,
|
| 199 |
+
"step": 29
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 20.0,
|
| 203 |
+
"learning_rate": 0.0001900968867902419,
|
| 204 |
+
"loss": 0.74,
|
| 205 |
+
"step": 30
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 20.0,
|
| 209 |
+
"eval_loss": 1.0189049243927002,
|
| 210 |
+
"eval_runtime": 0.8726,
|
| 211 |
+
"eval_samples_per_second": 6.876,
|
| 212 |
+
"eval_steps_per_second": 3.438,
|
| 213 |
+
"step": 30
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 21.0,
|
| 217 |
+
"learning_rate": 0.0001891006524188368,
|
| 218 |
+
"loss": 0.7203,
|
| 219 |
+
"step": 31
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 21.33,
|
| 223 |
+
"learning_rate": 0.0001880595531856738,
|
| 224 |
+
"loss": 0.2385,
|
| 225 |
+
"step": 32
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 22.0,
|
| 229 |
+
"learning_rate": 0.00018697411331556956,
|
| 230 |
+
"loss": 0.4841,
|
| 231 |
+
"step": 33
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 22.67,
|
| 235 |
+
"learning_rate": 0.00018584487936018661,
|
| 236 |
+
"loss": 0.4746,
|
| 237 |
+
"step": 34
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 23.0,
|
| 241 |
+
"learning_rate": 0.00018467241992282843,
|
| 242 |
+
"loss": 0.2257,
|
| 243 |
+
"step": 35
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 24.0,
|
| 247 |
+
"learning_rate": 0.00018345732537213027,
|
| 248 |
+
"loss": 0.6933,
|
| 249 |
+
"step": 36
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 25.0,
|
| 253 |
+
"learning_rate": 0.00018220020754479102,
|
| 254 |
+
"loss": 0.6819,
|
| 255 |
+
"step": 37
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 25.33,
|
| 259 |
+
"learning_rate": 0.00018090169943749476,
|
| 260 |
+
"loss": 0.2184,
|
| 261 |
+
"step": 38
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 26.0,
|
| 265 |
+
"learning_rate": 0.00017956245488817812,
|
| 266 |
+
"loss": 0.4415,
|
| 267 |
+
"step": 39
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 26.67,
|
| 271 |
+
"learning_rate": 0.000178183148246803,
|
| 272 |
+
"loss": 0.4336,
|
| 273 |
+
"step": 40
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 26.67,
|
| 277 |
+
"eval_loss": 0.9619529843330383,
|
| 278 |
+
"eval_runtime": 0.874,
|
| 279 |
+
"eval_samples_per_second": 6.865,
|
| 280 |
+
"eval_steps_per_second": 3.433,
|
| 281 |
+
"step": 40
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 27.0,
|
| 285 |
+
"learning_rate": 0.0001767644740358011,
|
| 286 |
+
"loss": 0.2187,
|
| 287 |
+
"step": 41
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 28.0,
|
| 291 |
+
"learning_rate": 0.00017530714660036112,
|
| 292 |
+
"loss": 0.642,
|
| 293 |
+
"step": 42
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 29.0,
|
| 297 |
+
"learning_rate": 0.00017381189974873407,
|
| 298 |
+
"loss": 0.6302,
|
| 299 |
+
"step": 43
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 29.33,
|
| 303 |
+
"learning_rate": 0.00017227948638273916,
|
| 304 |
+
"loss": 0.2054,
|
| 305 |
+
"step": 44
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 30.0,
|
| 309 |
+
"learning_rate": 0.00017071067811865476,
|
| 310 |
+
"loss": 0.4056,
|
| 311 |
+
"step": 45
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 30.67,
|
| 315 |
+
"learning_rate": 0.00016910626489868649,
|
| 316 |
+
"loss": 0.3933,
|
| 317 |
+
"step": 46
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 31.0,
|
| 321 |
+
"learning_rate": 0.00016746705459320745,
|
| 322 |
+
"loss": 0.1891,
|
| 323 |
+
"step": 47
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 32.0,
|
| 327 |
+
"learning_rate": 0.00016579387259397127,
|
| 328 |
+
"loss": 0.5689,
|
| 329 |
+
"step": 48
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 33.0,
|
| 333 |
+
"learning_rate": 0.0001640875613985024,
|
| 334 |
+
"loss": 0.5574,
|
| 335 |
+
"step": 49
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 33.33,
|
| 339 |
+
"learning_rate": 0.00016234898018587337,
|
| 340 |
+
"loss": 0.1827,
|
| 341 |
+
"step": 50
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 33.33,
|
| 345 |
+
"eval_loss": 0.8636454939842224,
|
| 346 |
+
"eval_runtime": 0.8737,
|
| 347 |
+
"eval_samples_per_second": 6.867,
|
| 348 |
+
"eval_steps_per_second": 3.434,
|
| 349 |
+
"step": 50
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 34.0,
|
| 353 |
+
"learning_rate": 0.000160579004384082,
|
| 354 |
+
"loss": 0.3592,
|
| 355 |
+
"step": 51
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 34.67,
|
| 359 |
+
"learning_rate": 0.00015877852522924732,
|
| 360 |
+
"loss": 0.3602,
|
| 361 |
+
"step": 52
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 35.0,
|
| 365 |
+
"learning_rate": 0.0001569484493168452,
|
| 366 |
+
"loss": 0.1678,
|
| 367 |
+
"step": 53
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 36.0,
|
| 371 |
+
"learning_rate": 0.00015508969814521025,
|
| 372 |
+
"loss": 0.5046,
|
| 373 |
+
"step": 54
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 37.0,
|
| 377 |
+
"learning_rate": 0.00015320320765153367,
|
| 378 |
+
"loss": 0.5016,
|
| 379 |
+
"step": 55
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 37.33,
|
| 383 |
+
"learning_rate": 0.00015128992774059063,
|
| 384 |
+
"loss": 0.1718,
|
| 385 |
+
"step": 56
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 38.0,
|
| 389 |
+
"learning_rate": 0.0001493508218064347,
|
| 390 |
+
"loss": 0.3245,
|
| 391 |
+
"step": 57
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 38.67,
|
| 395 |
+
"learning_rate": 0.00014738686624729986,
|
| 396 |
+
"loss": 0.3233,
|
| 397 |
+
"step": 58
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 39.0,
|
| 401 |
+
"learning_rate": 0.00014539904997395468,
|
| 402 |
+
"loss": 0.1573,
|
| 403 |
+
"step": 59
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 40.0,
|
| 407 |
+
"learning_rate": 0.00014338837391175582,
|
| 408 |
+
"loss": 0.4726,
|
| 409 |
+
"step": 60
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 40.0,
|
| 413 |
+
"eval_loss": 0.8675475120544434,
|
| 414 |
+
"eval_runtime": 0.8737,
|
| 415 |
+
"eval_samples_per_second": 6.867,
|
| 416 |
+
"eval_steps_per_second": 3.434,
|
| 417 |
+
"step": 60
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 41.0,
|
| 421 |
+
"learning_rate": 0.00014135585049665207,
|
| 422 |
+
"loss": 0.4665,
|
| 423 |
+
"step": 61
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 41.33,
|
| 427 |
+
"learning_rate": 0.00013930250316539238,
|
| 428 |
+
"loss": 0.1493,
|
| 429 |
+
"step": 62
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 42.0,
|
| 433 |
+
"learning_rate": 0.00013722936584019453,
|
| 434 |
+
"loss": 0.3025,
|
| 435 |
+
"step": 63
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 42.67,
|
| 439 |
+
"learning_rate": 0.0001351374824081343,
|
| 440 |
+
"loss": 0.2971,
|
| 441 |
+
"step": 64
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 43.0,
|
| 445 |
+
"learning_rate": 0.00013302790619551674,
|
| 446 |
+
"loss": 0.1348,
|
| 447 |
+
"step": 65
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 44.0,
|
| 451 |
+
"learning_rate": 0.00013090169943749476,
|
| 452 |
+
"loss": 0.4268,
|
| 453 |
+
"step": 66
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 45.0,
|
| 457 |
+
"learning_rate": 0.00012875993274320173,
|
| 458 |
+
"loss": 0.4176,
|
| 459 |
+
"step": 67
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 45.33,
|
| 463 |
+
"learning_rate": 0.00012660368455666752,
|
| 464 |
+
"loss": 0.1309,
|
| 465 |
+
"step": 68
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 46.0,
|
| 469 |
+
"learning_rate": 0.0001244340406137894,
|
| 470 |
+
"loss": 0.2693,
|
| 471 |
+
"step": 69
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 46.67,
|
| 475 |
+
"learning_rate": 0.00012225209339563145,
|
| 476 |
+
"loss": 0.2564,
|
| 477 |
+
"step": 70
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 46.67,
|
| 481 |
+
"eval_loss": 0.9176617860794067,
|
| 482 |
+
"eval_runtime": 0.8733,
|
| 483 |
+
"eval_samples_per_second": 6.871,
|
| 484 |
+
"eval_steps_per_second": 3.435,
|
| 485 |
+
"step": 70
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 47.0,
|
| 489 |
+
"learning_rate": 0.00012005894157832729,
|
| 490 |
+
"loss": 0.1313,
|
| 491 |
+
"step": 71
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 48.0,
|
| 495 |
+
"learning_rate": 0.00011785568947986367,
|
| 496 |
+
"loss": 0.3761,
|
| 497 |
+
"step": 72
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 49.0,
|
| 501 |
+
"learning_rate": 0.0001156434465040231,
|
| 502 |
+
"loss": 0.3719,
|
| 503 |
+
"step": 73
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 49.33,
|
| 507 |
+
"learning_rate": 0.00011342332658176555,
|
| 508 |
+
"loss": 0.1238,
|
| 509 |
+
"step": 74
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 50.0,
|
| 513 |
+
"learning_rate": 0.00011119644761033078,
|
| 514 |
+
"loss": 0.2303,
|
| 515 |
+
"step": 75
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 50.67,
|
| 519 |
+
"learning_rate": 0.00010896393089034336,
|
| 520 |
+
"loss": 0.2324,
|
| 521 |
+
"step": 76
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 51.0,
|
| 525 |
+
"learning_rate": 0.00010672690056120399,
|
| 526 |
+
"loss": 0.1158,
|
| 527 |
+
"step": 77
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 52.0,
|
| 531 |
+
"learning_rate": 0.00010448648303505151,
|
| 532 |
+
"loss": 0.3286,
|
| 533 |
+
"step": 78
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 53.0,
|
| 537 |
+
"learning_rate": 0.00010224380642958052,
|
| 538 |
+
"loss": 0.3256,
|
| 539 |
+
"step": 79
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 53.33,
|
| 543 |
+
"learning_rate": 0.0001,
|
| 544 |
+
"loss": 0.1031,
|
| 545 |
+
"step": 80
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 53.33,
|
| 549 |
+
"eval_loss": 0.9727662205696106,
|
| 550 |
+
"eval_runtime": 0.8715,
|
| 551 |
+
"eval_samples_per_second": 6.884,
|
| 552 |
+
"eval_steps_per_second": 3.442,
|
| 553 |
+
"step": 80
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 54.0,
|
| 557 |
+
"learning_rate": 9.775619357041952e-05,
|
| 558 |
+
"loss": 0.2059,
|
| 559 |
+
"step": 81
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 54.67,
|
| 563 |
+
"learning_rate": 9.551351696494854e-05,
|
| 564 |
+
"loss": 0.2032,
|
| 565 |
+
"step": 82
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 55.0,
|
| 569 |
+
"learning_rate": 9.327309943879604e-05,
|
| 570 |
+
"loss": 0.0924,
|
| 571 |
+
"step": 83
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 56.0,
|
| 575 |
+
"learning_rate": 9.103606910965666e-05,
|
| 576 |
+
"loss": 0.2809,
|
| 577 |
+
"step": 84
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 57.0,
|
| 581 |
+
"learning_rate": 8.880355238966923e-05,
|
| 582 |
+
"loss": 0.2757,
|
| 583 |
+
"step": 85
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 57.33,
|
| 587 |
+
"learning_rate": 8.657667341823448e-05,
|
| 588 |
+
"loss": 0.088,
|
| 589 |
+
"step": 86
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 58.0,
|
| 593 |
+
"learning_rate": 8.435655349597689e-05,
|
| 594 |
+
"loss": 0.1725,
|
| 595 |
+
"step": 87
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 58.67,
|
| 599 |
+
"learning_rate": 8.214431052013634e-05,
|
| 600 |
+
"loss": 0.1654,
|
| 601 |
+
"step": 88
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 59.0,
|
| 605 |
+
"learning_rate": 7.994105842167273e-05,
|
| 606 |
+
"loss": 0.0806,
|
| 607 |
+
"step": 89
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 60.0,
|
| 611 |
+
"learning_rate": 7.774790660436858e-05,
|
| 612 |
+
"loss": 0.2316,
|
| 613 |
+
"step": 90
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 60.0,
|
| 617 |
+
"eval_loss": 1.0534545183181763,
|
| 618 |
+
"eval_runtime": 0.8706,
|
| 619 |
+
"eval_samples_per_second": 6.892,
|
| 620 |
+
"eval_steps_per_second": 3.446,
|
| 621 |
+
"step": 90
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 61.0,
|
| 625 |
+
"learning_rate": 7.556595938621058e-05,
|
| 626 |
+
"loss": 0.2254,
|
| 627 |
+
"step": 91
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 61.33,
|
| 631 |
+
"learning_rate": 7.339631544333249e-05,
|
| 632 |
+
"loss": 0.0724,
|
| 633 |
+
"step": 92
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 62.0,
|
| 637 |
+
"learning_rate": 7.124006725679828e-05,
|
| 638 |
+
"loss": 0.1469,
|
| 639 |
+
"step": 93
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 62.67,
|
| 643 |
+
"learning_rate": 6.909830056250527e-05,
|
| 644 |
+
"loss": 0.1392,
|
| 645 |
+
"step": 94
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 63.0,
|
| 649 |
+
"learning_rate": 6.697209380448333e-05,
|
| 650 |
+
"loss": 0.0677,
|
| 651 |
+
"step": 95
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 64.0,
|
| 655 |
+
"learning_rate": 6.486251759186572e-05,
|
| 656 |
+
"loss": 0.1941,
|
| 657 |
+
"step": 96
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 65.0,
|
| 661 |
+
"learning_rate": 6.277063415980549e-05,
|
| 662 |
+
"loss": 0.1933,
|
| 663 |
+
"step": 97
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 65.33,
|
| 667 |
+
"learning_rate": 6.069749683460765e-05,
|
| 668 |
+
"loss": 0.0599,
|
| 669 |
+
"step": 98
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 66.0,
|
| 673 |
+
"learning_rate": 5.864414950334796e-05,
|
| 674 |
+
"loss": 0.1258,
|
| 675 |
+
"step": 99
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 66.67,
|
| 679 |
+
"learning_rate": 5.6611626088244194e-05,
|
| 680 |
+
"loss": 0.1146,
|
| 681 |
+
"step": 100
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 66.67,
|
| 685 |
+
"eval_loss": 1.147660493850708,
|
| 686 |
+
"eval_runtime": 0.8723,
|
| 687 |
+
"eval_samples_per_second": 6.878,
|
| 688 |
+
"eval_steps_per_second": 3.439,
|
| 689 |
+
"step": 100
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 67.0,
|
| 693 |
+
"learning_rate": 5.4600950026045326e-05,
|
| 694 |
+
"loss": 0.0585,
|
| 695 |
+
"step": 101
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 68.0,
|
| 699 |
+
"learning_rate": 5.261313375270014e-05,
|
| 700 |
+
"loss": 0.1722,
|
| 701 |
+
"step": 102
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 69.0,
|
| 705 |
+
"learning_rate": 5.0649178193565314e-05,
|
| 706 |
+
"loss": 0.1656,
|
| 707 |
+
"step": 103
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 69.33,
|
| 711 |
+
"learning_rate": 4.87100722594094e-05,
|
| 712 |
+
"loss": 0.0534,
|
| 713 |
+
"step": 104
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 70.0,
|
| 717 |
+
"learning_rate": 4.6796792348466356e-05,
|
| 718 |
+
"loss": 0.1112,
|
| 719 |
+
"step": 105
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 70.67,
|
| 723 |
+
"learning_rate": 4.491030185478976e-05,
|
| 724 |
+
"loss": 0.107,
|
| 725 |
+
"step": 106
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 71.0,
|
| 729 |
+
"learning_rate": 4.305155068315481e-05,
|
| 730 |
+
"loss": 0.0502,
|
| 731 |
+
"step": 107
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 72.0,
|
| 735 |
+
"learning_rate": 4.12214747707527e-05,
|
| 736 |
+
"loss": 0.1502,
|
| 737 |
+
"step": 108
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 73.0,
|
| 741 |
+
"learning_rate": 3.942099561591802e-05,
|
| 742 |
+
"loss": 0.1472,
|
| 743 |
+
"step": 109
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 73.33,
|
| 747 |
+
"learning_rate": 3.7651019814126654e-05,
|
| 748 |
+
"loss": 0.0463,
|
| 749 |
+
"step": 110
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 73.33,
|
| 753 |
+
"eval_loss": 1.2258118391036987,
|
| 754 |
+
"eval_runtime": 0.871,
|
| 755 |
+
"eval_samples_per_second": 6.888,
|
| 756 |
+
"eval_steps_per_second": 3.444,
|
| 757 |
+
"step": 110
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 74.0,
|
| 761 |
+
"learning_rate": 3.591243860149759e-05,
|
| 762 |
+
"loss": 0.0995,
|
| 763 |
+
"step": 111
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 74.67,
|
| 767 |
+
"learning_rate": 3.4206127406028745e-05,
|
| 768 |
+
"loss": 0.093,
|
| 769 |
+
"step": 112
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 75.0,
|
| 773 |
+
"learning_rate": 3.253294540679257e-05,
|
| 774 |
+
"loss": 0.0489,
|
| 775 |
+
"step": 113
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 76.0,
|
| 779 |
+
"learning_rate": 3.089373510131354e-05,
|
| 780 |
+
"loss": 0.1373,
|
| 781 |
+
"step": 114
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 77.0,
|
| 785 |
+
"learning_rate": 2.9289321881345254e-05,
|
| 786 |
+
"loss": 0.1337,
|
| 787 |
+
"step": 115
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 77.33,
|
| 791 |
+
"learning_rate": 2.7720513617260856e-05,
|
| 792 |
+
"loss": 0.0413,
|
| 793 |
+
"step": 116
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 78.0,
|
| 797 |
+
"learning_rate": 2.6188100251265945e-05,
|
| 798 |
+
"loss": 0.0911,
|
| 799 |
+
"step": 117
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 78.67,
|
| 803 |
+
"learning_rate": 2.4692853399638917e-05,
|
| 804 |
+
"loss": 0.0871,
|
| 805 |
+
"step": 118
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 79.0,
|
| 809 |
+
"learning_rate": 2.323552596419889e-05,
|
| 810 |
+
"loss": 0.0424,
|
| 811 |
+
"step": 119
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 80.0,
|
| 815 |
+
"learning_rate": 2.181685175319702e-05,
|
| 816 |
+
"loss": 0.127,
|
| 817 |
+
"step": 120
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 80.0,
|
| 821 |
+
"eval_loss": 1.2826638221740723,
|
| 822 |
+
"eval_runtime": 0.8716,
|
| 823 |
+
"eval_samples_per_second": 6.884,
|
| 824 |
+
"eval_steps_per_second": 3.442,
|
| 825 |
+
"step": 120
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 81.0,
|
| 829 |
+
"learning_rate": 2.043754511182191e-05,
|
| 830 |
+
"loss": 0.124,
|
| 831 |
+
"step": 121
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 81.33,
|
| 835 |
+
"learning_rate": 1.9098300562505266e-05,
|
| 836 |
+
"loss": 0.0409,
|
| 837 |
+
"step": 122
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 82.0,
|
| 841 |
+
"learning_rate": 1.7799792455209018e-05,
|
| 842 |
+
"loss": 0.0819,
|
| 843 |
+
"step": 123
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 82.67,
|
| 847 |
+
"learning_rate": 1.6542674627869737e-05,
|
| 848 |
+
"loss": 0.0816,
|
| 849 |
+
"step": 124
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 83.0,
|
| 853 |
+
"learning_rate": 1.5327580077171587e-05,
|
| 854 |
+
"loss": 0.0397,
|
| 855 |
+
"step": 125
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 84.0,
|
| 859 |
+
"learning_rate": 1.415512063981339e-05,
|
| 860 |
+
"loss": 0.1216,
|
| 861 |
+
"step": 126
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 85.0,
|
| 865 |
+
"learning_rate": 1.3025886684430467e-05,
|
| 866 |
+
"loss": 0.1204,
|
| 867 |
+
"step": 127
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 85.33,
|
| 871 |
+
"learning_rate": 1.19404468143262e-05,
|
| 872 |
+
"loss": 0.041,
|
| 873 |
+
"step": 128
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 86.0,
|
| 877 |
+
"learning_rate": 1.0899347581163221e-05,
|
| 878 |
+
"loss": 0.0772,
|
| 879 |
+
"step": 129
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 86.67,
|
| 883 |
+
"learning_rate": 9.903113209758096e-06,
|
| 884 |
+
"loss": 0.0779,
|
| 885 |
+
"step": 130
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 86.67,
|
| 889 |
+
"eval_loss": 1.3157517910003662,
|
| 890 |
+
"eval_runtime": 0.8712,
|
| 891 |
+
"eval_samples_per_second": 6.887,
|
| 892 |
+
"eval_steps_per_second": 3.444,
|
| 893 |
+
"step": 130
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 87.0,
|
| 897 |
+
"learning_rate": 8.952245334118414e-06,
|
| 898 |
+
"loss": 0.0401,
|
| 899 |
+
"step": 131
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 88.0,
|
| 903 |
+
"learning_rate": 8.047222744854943e-06,
|
| 904 |
+
"loss": 0.1173,
|
| 905 |
+
"step": 132
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 89.0,
|
| 909 |
+
"learning_rate": 7.1885011480961164e-06,
|
| 910 |
+
"loss": 0.1178,
|
| 911 |
+
"step": 133
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 89.33,
|
| 915 |
+
"learning_rate": 6.37651293602628e-06,
|
| 916 |
+
"loss": 0.0397,
|
| 917 |
+
"step": 134
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 90.0,
|
| 921 |
+
"learning_rate": 5.611666969163243e-06,
|
| 922 |
+
"loss": 0.079,
|
| 923 |
+
"step": 135
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 90.67,
|
| 927 |
+
"learning_rate": 4.8943483704846475e-06,
|
| 928 |
+
"loss": 0.0766,
|
| 929 |
+
"step": 136
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 91.0,
|
| 933 |
+
"learning_rate": 4.224918331506955e-06,
|
| 934 |
+
"loss": 0.0397,
|
| 935 |
+
"step": 137
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 92.0,
|
| 939 |
+
"learning_rate": 3.6037139304146762e-06,
|
| 940 |
+
"loss": 0.1141,
|
| 941 |
+
"step": 138
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 93.0,
|
| 945 |
+
"learning_rate": 3.0310479623313127e-06,
|
| 946 |
+
"loss": 0.1184,
|
| 947 |
+
"step": 139
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 93.33,
|
| 951 |
+
"learning_rate": 2.5072087818176382e-06,
|
| 952 |
+
"loss": 0.0386,
|
| 953 |
+
"step": 140
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 93.33,
|
| 957 |
+
"eval_loss": 1.331322431564331,
|
| 958 |
+
"eval_runtime": 0.873,
|
| 959 |
+
"eval_samples_per_second": 6.873,
|
| 960 |
+
"eval_steps_per_second": 3.436,
|
| 961 |
+
"step": 140
|
| 962 |
+
}
|
| 963 |
+
],
|
| 964 |
+
"logging_steps": 1,
|
| 965 |
+
"max_steps": 150,
|
| 966 |
+
"num_train_epochs": 150,
|
| 967 |
+
"save_steps": 20,
|
| 968 |
+
"total_flos": 9.203523552018432e+16,
|
| 969 |
+
"trial_name": null,
|
| 970 |
+
"trial_params": null
|
| 971 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bb32c45a82deea548f0ed3efa27ff0394baec1e7710f8c61760f66cfe630d33c
|
| 3 |
+
size 4475
|