Farouk commited on
Commit ·
09cb6dc
1
Parent(s): 40d2def
Training in progress, step 6800
Browse files- adapter_model.bin +1 -1
- checkpoint-3400/adapter_model/adapter_model/README.md +12 -0
- checkpoint-3400/adapter_model/adapter_model/adapter_model.bin +1 -1
- {checkpoint-4800 → checkpoint-6800}/README.md +0 -0
- {checkpoint-4800 → checkpoint-6800}/adapter_config.json +0 -0
- {checkpoint-4800 → checkpoint-6800}/adapter_model.bin +1 -1
- {checkpoint-4800 → checkpoint-6800}/added_tokens.json +0 -0
- {checkpoint-4800 → checkpoint-6800}/optimizer.pt +1 -1
- {checkpoint-4800 → checkpoint-6800}/rng_state.pth +1 -1
- {checkpoint-4800 → checkpoint-6800}/scheduler.pt +1 -1
- {checkpoint-4800 → checkpoint-6800}/special_tokens_map.json +0 -0
- {checkpoint-4800 → checkpoint-6800}/tokenizer.model +0 -0
- {checkpoint-4800 → checkpoint-6800}/tokenizer_config.json +0 -0
- {checkpoint-4800 → checkpoint-6800}/trainer_state.json +1913 -3
- {checkpoint-4800 → checkpoint-6800}/training_args.bin +0 -0
adapter_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 319977229
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:644bbce560a68e10fbb77f0f9b0b25fea68fef7441125e052271f0ab3cc41fa5
|
| 3 |
size 319977229
|
checkpoint-3400/adapter_model/adapter_model/README.md
CHANGED
|
@@ -169,6 +169,17 @@ The following `bitsandbytes` quantization config was used during training:
|
|
| 169 |
- bnb_4bit_use_double_quant: True
|
| 170 |
- bnb_4bit_compute_dtype: bfloat16
|
| 171 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 172 |
The following `bitsandbytes` quantization config was used during training:
|
| 173 |
- load_in_8bit: False
|
| 174 |
- load_in_4bit: True
|
|
@@ -196,5 +207,6 @@ The following `bitsandbytes` quantization config was used during training:
|
|
| 196 |
- PEFT 0.4.0
|
| 197 |
- PEFT 0.4.0
|
| 198 |
- PEFT 0.4.0
|
|
|
|
| 199 |
|
| 200 |
- PEFT 0.4.0
|
|
|
|
| 169 |
- bnb_4bit_use_double_quant: True
|
| 170 |
- bnb_4bit_compute_dtype: bfloat16
|
| 171 |
|
| 172 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 173 |
+
- load_in_8bit: False
|
| 174 |
+
- load_in_4bit: True
|
| 175 |
+
- llm_int8_threshold: 6.0
|
| 176 |
+
- llm_int8_skip_modules: None
|
| 177 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 178 |
+
- llm_int8_has_fp16_weight: False
|
| 179 |
+
- bnb_4bit_quant_type: nf4
|
| 180 |
+
- bnb_4bit_use_double_quant: True
|
| 181 |
+
- bnb_4bit_compute_dtype: bfloat16
|
| 182 |
+
|
| 183 |
The following `bitsandbytes` quantization config was used during training:
|
| 184 |
- load_in_8bit: False
|
| 185 |
- load_in_4bit: True
|
|
|
|
| 207 |
- PEFT 0.4.0
|
| 208 |
- PEFT 0.4.0
|
| 209 |
- PEFT 0.4.0
|
| 210 |
+
- PEFT 0.4.0
|
| 211 |
|
| 212 |
- PEFT 0.4.0
|
checkpoint-3400/adapter_model/adapter_model/adapter_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 319977229
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8d21530bb4f31e884eaf88f3e0cd06e11d12d59f60e425c58c610d77c80cb4bc
|
| 3 |
size 319977229
|
{checkpoint-4800 → checkpoint-6800}/README.md
RENAMED
|
File without changes
|
{checkpoint-4800 → checkpoint-6800}/adapter_config.json
RENAMED
|
File without changes
|
{checkpoint-4800 → checkpoint-6800}/adapter_model.bin
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 319977229
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:644bbce560a68e10fbb77f0f9b0b25fea68fef7441125e052271f0ab3cc41fa5
|
| 3 |
size 319977229
|
{checkpoint-4800 → checkpoint-6800}/added_tokens.json
RENAMED
|
File without changes
|
{checkpoint-4800 → checkpoint-6800}/optimizer.pt
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1279539973
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a426ca015d8b697d3a434aee5599a63dea291bcf50251e40ce08e69d8ce3476c
|
| 3 |
size 1279539973
|
{checkpoint-4800 → checkpoint-6800}/rng_state.pth
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14511
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ab39d01711b16fd9420c23084db864739c6a6a588204bb41f4b799ca01da87fd
|
| 3 |
size 14511
|
{checkpoint-4800 → checkpoint-6800}/scheduler.pt
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:97c7e7304629a0d5b32d77774ba8769b4783d1dab8e735fe47e12df568c740c2
|
| 3 |
size 627
|
{checkpoint-4800 → checkpoint-6800}/special_tokens_map.json
RENAMED
|
File without changes
|
{checkpoint-4800 → checkpoint-6800}/tokenizer.model
RENAMED
|
File without changes
|
{checkpoint-4800 → checkpoint-6800}/tokenizer_config.json
RENAMED
|
File without changes
|
{checkpoint-4800 → checkpoint-6800}/trainer_state.json
RENAMED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": 0.7174572944641113,
|
| 3 |
"best_model_checkpoint": "experts/expert-30/checkpoint-3400",
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -4590,11 +4590,1921 @@
|
|
| 4590 |
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 4591 |
"mmlu_loss": 1.5985829301820411,
|
| 4592 |
"step": 4800
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4593 |
}
|
| 4594 |
],
|
| 4595 |
"max_steps": 10000,
|
| 4596 |
"num_train_epochs": 6,
|
| 4597 |
-
"total_flos":
|
| 4598 |
"trial_name": null,
|
| 4599 |
"trial_params": null
|
| 4600 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": 0.7174572944641113,
|
| 3 |
"best_model_checkpoint": "experts/expert-30/checkpoint-3400",
|
| 4 |
+
"epoch": 3.784084585420145,
|
| 5 |
+
"global_step": 6800,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 4590 |
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 4591 |
"mmlu_loss": 1.5985829301820411,
|
| 4592 |
"step": 4800
|
| 4593 |
+
},
|
| 4594 |
+
{
|
| 4595 |
+
"epoch": 2.68,
|
| 4596 |
+
"learning_rate": 0.0002,
|
| 4597 |
+
"loss": 0.5994,
|
| 4598 |
+
"step": 4810
|
| 4599 |
+
},
|
| 4600 |
+
{
|
| 4601 |
+
"epoch": 2.68,
|
| 4602 |
+
"learning_rate": 0.0002,
|
| 4603 |
+
"loss": 0.6289,
|
| 4604 |
+
"step": 4820
|
| 4605 |
+
},
|
| 4606 |
+
{
|
| 4607 |
+
"epoch": 2.69,
|
| 4608 |
+
"learning_rate": 0.0002,
|
| 4609 |
+
"loss": 0.6103,
|
| 4610 |
+
"step": 4830
|
| 4611 |
+
},
|
| 4612 |
+
{
|
| 4613 |
+
"epoch": 2.69,
|
| 4614 |
+
"learning_rate": 0.0002,
|
| 4615 |
+
"loss": 0.566,
|
| 4616 |
+
"step": 4840
|
| 4617 |
+
},
|
| 4618 |
+
{
|
| 4619 |
+
"epoch": 2.7,
|
| 4620 |
+
"learning_rate": 0.0002,
|
| 4621 |
+
"loss": 0.5673,
|
| 4622 |
+
"step": 4850
|
| 4623 |
+
},
|
| 4624 |
+
{
|
| 4625 |
+
"epoch": 2.7,
|
| 4626 |
+
"learning_rate": 0.0002,
|
| 4627 |
+
"loss": 0.6362,
|
| 4628 |
+
"step": 4860
|
| 4629 |
+
},
|
| 4630 |
+
{
|
| 4631 |
+
"epoch": 2.71,
|
| 4632 |
+
"learning_rate": 0.0002,
|
| 4633 |
+
"loss": 0.6337,
|
| 4634 |
+
"step": 4870
|
| 4635 |
+
},
|
| 4636 |
+
{
|
| 4637 |
+
"epoch": 2.72,
|
| 4638 |
+
"learning_rate": 0.0002,
|
| 4639 |
+
"loss": 0.5901,
|
| 4640 |
+
"step": 4880
|
| 4641 |
+
},
|
| 4642 |
+
{
|
| 4643 |
+
"epoch": 2.72,
|
| 4644 |
+
"learning_rate": 0.0002,
|
| 4645 |
+
"loss": 0.6435,
|
| 4646 |
+
"step": 4890
|
| 4647 |
+
},
|
| 4648 |
+
{
|
| 4649 |
+
"epoch": 2.73,
|
| 4650 |
+
"learning_rate": 0.0002,
|
| 4651 |
+
"loss": 0.5337,
|
| 4652 |
+
"step": 4900
|
| 4653 |
+
},
|
| 4654 |
+
{
|
| 4655 |
+
"epoch": 2.73,
|
| 4656 |
+
"learning_rate": 0.0002,
|
| 4657 |
+
"loss": 0.5591,
|
| 4658 |
+
"step": 4910
|
| 4659 |
+
},
|
| 4660 |
+
{
|
| 4661 |
+
"epoch": 2.74,
|
| 4662 |
+
"learning_rate": 0.0002,
|
| 4663 |
+
"loss": 0.5847,
|
| 4664 |
+
"step": 4920
|
| 4665 |
+
},
|
| 4666 |
+
{
|
| 4667 |
+
"epoch": 2.74,
|
| 4668 |
+
"learning_rate": 0.0002,
|
| 4669 |
+
"loss": 0.6051,
|
| 4670 |
+
"step": 4930
|
| 4671 |
+
},
|
| 4672 |
+
{
|
| 4673 |
+
"epoch": 2.75,
|
| 4674 |
+
"learning_rate": 0.0002,
|
| 4675 |
+
"loss": 0.5884,
|
| 4676 |
+
"step": 4940
|
| 4677 |
+
},
|
| 4678 |
+
{
|
| 4679 |
+
"epoch": 2.75,
|
| 4680 |
+
"learning_rate": 0.0002,
|
| 4681 |
+
"loss": 0.5731,
|
| 4682 |
+
"step": 4950
|
| 4683 |
+
},
|
| 4684 |
+
{
|
| 4685 |
+
"epoch": 2.76,
|
| 4686 |
+
"learning_rate": 0.0002,
|
| 4687 |
+
"loss": 0.5878,
|
| 4688 |
+
"step": 4960
|
| 4689 |
+
},
|
| 4690 |
+
{
|
| 4691 |
+
"epoch": 2.77,
|
| 4692 |
+
"learning_rate": 0.0002,
|
| 4693 |
+
"loss": 0.6194,
|
| 4694 |
+
"step": 4970
|
| 4695 |
+
},
|
| 4696 |
+
{
|
| 4697 |
+
"epoch": 2.77,
|
| 4698 |
+
"learning_rate": 0.0002,
|
| 4699 |
+
"loss": 0.5798,
|
| 4700 |
+
"step": 4980
|
| 4701 |
+
},
|
| 4702 |
+
{
|
| 4703 |
+
"epoch": 2.78,
|
| 4704 |
+
"learning_rate": 0.0002,
|
| 4705 |
+
"loss": 0.5841,
|
| 4706 |
+
"step": 4990
|
| 4707 |
+
},
|
| 4708 |
+
{
|
| 4709 |
+
"epoch": 2.78,
|
| 4710 |
+
"learning_rate": 0.0002,
|
| 4711 |
+
"loss": 0.61,
|
| 4712 |
+
"step": 5000
|
| 4713 |
+
},
|
| 4714 |
+
{
|
| 4715 |
+
"epoch": 2.78,
|
| 4716 |
+
"eval_loss": 0.7293540239334106,
|
| 4717 |
+
"eval_runtime": 80.9574,
|
| 4718 |
+
"eval_samples_per_second": 12.352,
|
| 4719 |
+
"eval_steps_per_second": 6.176,
|
| 4720 |
+
"step": 5000
|
| 4721 |
+
},
|
| 4722 |
+
{
|
| 4723 |
+
"epoch": 2.78,
|
| 4724 |
+
"mmlu_eval_accuracy": 0.48996341553722794,
|
| 4725 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
|
| 4726 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 4727 |
+
"mmlu_eval_accuracy_astronomy": 0.5625,
|
| 4728 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 4729 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
|
| 4730 |
+
"mmlu_eval_accuracy_college_biology": 0.5,
|
| 4731 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 4732 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 4733 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 4734 |
+
"mmlu_eval_accuracy_college_medicine": 0.36363636363636365,
|
| 4735 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 4736 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 4737 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 4738 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 4739 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.25,
|
| 4740 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
|
| 4741 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 4742 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 4743 |
+
"mmlu_eval_accuracy_high_school_biology": 0.40625,
|
| 4744 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
|
| 4745 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 4746 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
|
| 4747 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
|
| 4748 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
|
| 4749 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4418604651162791,
|
| 4750 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
|
| 4751 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
|
| 4752 |
+
"mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
|
| 4753 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
|
| 4754 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913,
|
| 4755 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
|
| 4756 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
|
| 4757 |
+
"mmlu_eval_accuracy_human_aging": 0.6086956521739131,
|
| 4758 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 4759 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 4760 |
+
"mmlu_eval_accuracy_jurisprudence": 0.2727272727272727,
|
| 4761 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
|
| 4762 |
+
"mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
|
| 4763 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 4764 |
+
"mmlu_eval_accuracy_marketing": 0.8,
|
| 4765 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 4766 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
|
| 4767 |
+
"mmlu_eval_accuracy_moral_disputes": 0.42105263157894735,
|
| 4768 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.28,
|
| 4769 |
+
"mmlu_eval_accuracy_nutrition": 0.7575757575757576,
|
| 4770 |
+
"mmlu_eval_accuracy_philosophy": 0.5,
|
| 4771 |
+
"mmlu_eval_accuracy_prehistory": 0.4857142857142857,
|
| 4772 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
|
| 4773 |
+
"mmlu_eval_accuracy_professional_law": 0.3235294117647059,
|
| 4774 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
|
| 4775 |
+
"mmlu_eval_accuracy_professional_psychology": 0.42028985507246375,
|
| 4776 |
+
"mmlu_eval_accuracy_public_relations": 0.5833333333333334,
|
| 4777 |
+
"mmlu_eval_accuracy_security_studies": 0.4444444444444444,
|
| 4778 |
+
"mmlu_eval_accuracy_sociology": 0.7727272727272727,
|
| 4779 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 4780 |
+
"mmlu_eval_accuracy_virology": 0.4444444444444444,
|
| 4781 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 4782 |
+
"mmlu_loss": 1.7054601166173622,
|
| 4783 |
+
"step": 5000
|
| 4784 |
+
},
|
| 4785 |
+
{
|
| 4786 |
+
"epoch": 2.79,
|
| 4787 |
+
"learning_rate": 0.0002,
|
| 4788 |
+
"loss": 0.6738,
|
| 4789 |
+
"step": 5010
|
| 4790 |
+
},
|
| 4791 |
+
{
|
| 4792 |
+
"epoch": 2.79,
|
| 4793 |
+
"learning_rate": 0.0002,
|
| 4794 |
+
"loss": 0.5561,
|
| 4795 |
+
"step": 5020
|
| 4796 |
+
},
|
| 4797 |
+
{
|
| 4798 |
+
"epoch": 2.8,
|
| 4799 |
+
"learning_rate": 0.0002,
|
| 4800 |
+
"loss": 0.6304,
|
| 4801 |
+
"step": 5030
|
| 4802 |
+
},
|
| 4803 |
+
{
|
| 4804 |
+
"epoch": 2.8,
|
| 4805 |
+
"learning_rate": 0.0002,
|
| 4806 |
+
"loss": 0.6727,
|
| 4807 |
+
"step": 5040
|
| 4808 |
+
},
|
| 4809 |
+
{
|
| 4810 |
+
"epoch": 2.81,
|
| 4811 |
+
"learning_rate": 0.0002,
|
| 4812 |
+
"loss": 0.6002,
|
| 4813 |
+
"step": 5050
|
| 4814 |
+
},
|
| 4815 |
+
{
|
| 4816 |
+
"epoch": 2.82,
|
| 4817 |
+
"learning_rate": 0.0002,
|
| 4818 |
+
"loss": 0.5528,
|
| 4819 |
+
"step": 5060
|
| 4820 |
+
},
|
| 4821 |
+
{
|
| 4822 |
+
"epoch": 2.82,
|
| 4823 |
+
"learning_rate": 0.0002,
|
| 4824 |
+
"loss": 0.6459,
|
| 4825 |
+
"step": 5070
|
| 4826 |
+
},
|
| 4827 |
+
{
|
| 4828 |
+
"epoch": 2.83,
|
| 4829 |
+
"learning_rate": 0.0002,
|
| 4830 |
+
"loss": 0.574,
|
| 4831 |
+
"step": 5080
|
| 4832 |
+
},
|
| 4833 |
+
{
|
| 4834 |
+
"epoch": 2.83,
|
| 4835 |
+
"learning_rate": 0.0002,
|
| 4836 |
+
"loss": 0.677,
|
| 4837 |
+
"step": 5090
|
| 4838 |
+
},
|
| 4839 |
+
{
|
| 4840 |
+
"epoch": 2.84,
|
| 4841 |
+
"learning_rate": 0.0002,
|
| 4842 |
+
"loss": 0.593,
|
| 4843 |
+
"step": 5100
|
| 4844 |
+
},
|
| 4845 |
+
{
|
| 4846 |
+
"epoch": 2.84,
|
| 4847 |
+
"learning_rate": 0.0002,
|
| 4848 |
+
"loss": 0.6833,
|
| 4849 |
+
"step": 5110
|
| 4850 |
+
},
|
| 4851 |
+
{
|
| 4852 |
+
"epoch": 2.85,
|
| 4853 |
+
"learning_rate": 0.0002,
|
| 4854 |
+
"loss": 0.6099,
|
| 4855 |
+
"step": 5120
|
| 4856 |
+
},
|
| 4857 |
+
{
|
| 4858 |
+
"epoch": 2.85,
|
| 4859 |
+
"learning_rate": 0.0002,
|
| 4860 |
+
"loss": 0.6767,
|
| 4861 |
+
"step": 5130
|
| 4862 |
+
},
|
| 4863 |
+
{
|
| 4864 |
+
"epoch": 2.86,
|
| 4865 |
+
"learning_rate": 0.0002,
|
| 4866 |
+
"loss": 0.5795,
|
| 4867 |
+
"step": 5140
|
| 4868 |
+
},
|
| 4869 |
+
{
|
| 4870 |
+
"epoch": 2.87,
|
| 4871 |
+
"learning_rate": 0.0002,
|
| 4872 |
+
"loss": 0.5887,
|
| 4873 |
+
"step": 5150
|
| 4874 |
+
},
|
| 4875 |
+
{
|
| 4876 |
+
"epoch": 2.87,
|
| 4877 |
+
"learning_rate": 0.0002,
|
| 4878 |
+
"loss": 0.6044,
|
| 4879 |
+
"step": 5160
|
| 4880 |
+
},
|
| 4881 |
+
{
|
| 4882 |
+
"epoch": 2.88,
|
| 4883 |
+
"learning_rate": 0.0002,
|
| 4884 |
+
"loss": 0.6168,
|
| 4885 |
+
"step": 5170
|
| 4886 |
+
},
|
| 4887 |
+
{
|
| 4888 |
+
"epoch": 2.88,
|
| 4889 |
+
"learning_rate": 0.0002,
|
| 4890 |
+
"loss": 0.6432,
|
| 4891 |
+
"step": 5180
|
| 4892 |
+
},
|
| 4893 |
+
{
|
| 4894 |
+
"epoch": 2.89,
|
| 4895 |
+
"learning_rate": 0.0002,
|
| 4896 |
+
"loss": 0.5827,
|
| 4897 |
+
"step": 5190
|
| 4898 |
+
},
|
| 4899 |
+
{
|
| 4900 |
+
"epoch": 2.89,
|
| 4901 |
+
"learning_rate": 0.0002,
|
| 4902 |
+
"loss": 0.6037,
|
| 4903 |
+
"step": 5200
|
| 4904 |
+
},
|
| 4905 |
+
{
|
| 4906 |
+
"epoch": 2.89,
|
| 4907 |
+
"eval_loss": 0.7285625338554382,
|
| 4908 |
+
"eval_runtime": 80.9058,
|
| 4909 |
+
"eval_samples_per_second": 12.36,
|
| 4910 |
+
"eval_steps_per_second": 6.18,
|
| 4911 |
+
"step": 5200
|
| 4912 |
+
},
|
| 4913 |
+
{
|
| 4914 |
+
"epoch": 2.89,
|
| 4915 |
+
"mmlu_eval_accuracy": 0.4848638106968446,
|
| 4916 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
|
| 4917 |
+
"mmlu_eval_accuracy_anatomy": 0.5714285714285714,
|
| 4918 |
+
"mmlu_eval_accuracy_astronomy": 0.5,
|
| 4919 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 4920 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
|
| 4921 |
+
"mmlu_eval_accuracy_college_biology": 0.5,
|
| 4922 |
+
"mmlu_eval_accuracy_college_chemistry": 0.375,
|
| 4923 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 4924 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 4925 |
+
"mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
|
| 4926 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 4927 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 4928 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 4929 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 4930 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.25,
|
| 4931 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.2926829268292683,
|
| 4932 |
+
"mmlu_eval_accuracy_formal_logic": 0.07142857142857142,
|
| 4933 |
+
"mmlu_eval_accuracy_global_facts": 0.4,
|
| 4934 |
+
"mmlu_eval_accuracy_high_school_biology": 0.375,
|
| 4935 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
|
| 4936 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 4937 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
|
| 4938 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
|
| 4939 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
|
| 4940 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4186046511627907,
|
| 4941 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
|
| 4942 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
|
| 4943 |
+
"mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
|
| 4944 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
|
| 4945 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913,
|
| 4946 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
|
| 4947 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
|
| 4948 |
+
"mmlu_eval_accuracy_human_aging": 0.6086956521739131,
|
| 4949 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 4950 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 4951 |
+
"mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
|
| 4952 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
|
| 4953 |
+
"mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
|
| 4954 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 4955 |
+
"mmlu_eval_accuracy_marketing": 0.76,
|
| 4956 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 4957 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6511627906976745,
|
| 4958 |
+
"mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
|
| 4959 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.29,
|
| 4960 |
+
"mmlu_eval_accuracy_nutrition": 0.7575757575757576,
|
| 4961 |
+
"mmlu_eval_accuracy_philosophy": 0.5,
|
| 4962 |
+
"mmlu_eval_accuracy_prehistory": 0.4857142857142857,
|
| 4963 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
|
| 4964 |
+
"mmlu_eval_accuracy_professional_law": 0.3352941176470588,
|
| 4965 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
|
| 4966 |
+
"mmlu_eval_accuracy_professional_psychology": 0.391304347826087,
|
| 4967 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 4968 |
+
"mmlu_eval_accuracy_security_studies": 0.4444444444444444,
|
| 4969 |
+
"mmlu_eval_accuracy_sociology": 0.7727272727272727,
|
| 4970 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 4971 |
+
"mmlu_eval_accuracy_virology": 0.4444444444444444,
|
| 4972 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 4973 |
+
"mmlu_loss": 1.5992514780545048,
|
| 4974 |
+
"step": 5200
|
| 4975 |
+
},
|
| 4976 |
+
{
|
| 4977 |
+
"epoch": 2.9,
|
| 4978 |
+
"learning_rate": 0.0002,
|
| 4979 |
+
"loss": 0.5656,
|
| 4980 |
+
"step": 5210
|
| 4981 |
+
},
|
| 4982 |
+
{
|
| 4983 |
+
"epoch": 2.9,
|
| 4984 |
+
"learning_rate": 0.0002,
|
| 4985 |
+
"loss": 0.5785,
|
| 4986 |
+
"step": 5220
|
| 4987 |
+
},
|
| 4988 |
+
{
|
| 4989 |
+
"epoch": 2.91,
|
| 4990 |
+
"learning_rate": 0.0002,
|
| 4991 |
+
"loss": 0.588,
|
| 4992 |
+
"step": 5230
|
| 4993 |
+
},
|
| 4994 |
+
{
|
| 4995 |
+
"epoch": 2.92,
|
| 4996 |
+
"learning_rate": 0.0002,
|
| 4997 |
+
"loss": 0.5189,
|
| 4998 |
+
"step": 5240
|
| 4999 |
+
},
|
| 5000 |
+
{
|
| 5001 |
+
"epoch": 2.92,
|
| 5002 |
+
"learning_rate": 0.0002,
|
| 5003 |
+
"loss": 0.6122,
|
| 5004 |
+
"step": 5250
|
| 5005 |
+
},
|
| 5006 |
+
{
|
| 5007 |
+
"epoch": 2.93,
|
| 5008 |
+
"learning_rate": 0.0002,
|
| 5009 |
+
"loss": 0.5854,
|
| 5010 |
+
"step": 5260
|
| 5011 |
+
},
|
| 5012 |
+
{
|
| 5013 |
+
"epoch": 2.93,
|
| 5014 |
+
"learning_rate": 0.0002,
|
| 5015 |
+
"loss": 0.6181,
|
| 5016 |
+
"step": 5270
|
| 5017 |
+
},
|
| 5018 |
+
{
|
| 5019 |
+
"epoch": 2.94,
|
| 5020 |
+
"learning_rate": 0.0002,
|
| 5021 |
+
"loss": 0.5908,
|
| 5022 |
+
"step": 5280
|
| 5023 |
+
},
|
| 5024 |
+
{
|
| 5025 |
+
"epoch": 2.94,
|
| 5026 |
+
"learning_rate": 0.0002,
|
| 5027 |
+
"loss": 0.6593,
|
| 5028 |
+
"step": 5290
|
| 5029 |
+
},
|
| 5030 |
+
{
|
| 5031 |
+
"epoch": 2.95,
|
| 5032 |
+
"learning_rate": 0.0002,
|
| 5033 |
+
"loss": 0.6073,
|
| 5034 |
+
"step": 5300
|
| 5035 |
+
},
|
| 5036 |
+
{
|
| 5037 |
+
"epoch": 2.95,
|
| 5038 |
+
"learning_rate": 0.0002,
|
| 5039 |
+
"loss": 0.6189,
|
| 5040 |
+
"step": 5310
|
| 5041 |
+
},
|
| 5042 |
+
{
|
| 5043 |
+
"epoch": 2.96,
|
| 5044 |
+
"learning_rate": 0.0002,
|
| 5045 |
+
"loss": 0.5978,
|
| 5046 |
+
"step": 5320
|
| 5047 |
+
},
|
| 5048 |
+
{
|
| 5049 |
+
"epoch": 2.97,
|
| 5050 |
+
"learning_rate": 0.0002,
|
| 5051 |
+
"loss": 0.5963,
|
| 5052 |
+
"step": 5330
|
| 5053 |
+
},
|
| 5054 |
+
{
|
| 5055 |
+
"epoch": 2.97,
|
| 5056 |
+
"learning_rate": 0.0002,
|
| 5057 |
+
"loss": 0.5567,
|
| 5058 |
+
"step": 5340
|
| 5059 |
+
},
|
| 5060 |
+
{
|
| 5061 |
+
"epoch": 2.98,
|
| 5062 |
+
"learning_rate": 0.0002,
|
| 5063 |
+
"loss": 0.5911,
|
| 5064 |
+
"step": 5350
|
| 5065 |
+
},
|
| 5066 |
+
{
|
| 5067 |
+
"epoch": 2.98,
|
| 5068 |
+
"learning_rate": 0.0002,
|
| 5069 |
+
"loss": 0.5642,
|
| 5070 |
+
"step": 5360
|
| 5071 |
+
},
|
| 5072 |
+
{
|
| 5073 |
+
"epoch": 2.99,
|
| 5074 |
+
"learning_rate": 0.0002,
|
| 5075 |
+
"loss": 0.6401,
|
| 5076 |
+
"step": 5370
|
| 5077 |
+
},
|
| 5078 |
+
{
|
| 5079 |
+
"epoch": 2.99,
|
| 5080 |
+
"learning_rate": 0.0002,
|
| 5081 |
+
"loss": 0.6885,
|
| 5082 |
+
"step": 5380
|
| 5083 |
+
},
|
| 5084 |
+
{
|
| 5085 |
+
"epoch": 3.0,
|
| 5086 |
+
"learning_rate": 0.0002,
|
| 5087 |
+
"loss": 0.5979,
|
| 5088 |
+
"step": 5390
|
| 5089 |
+
},
|
| 5090 |
+
{
|
| 5091 |
+
"epoch": 3.01,
|
| 5092 |
+
"learning_rate": 0.0002,
|
| 5093 |
+
"loss": 0.5147,
|
| 5094 |
+
"step": 5400
|
| 5095 |
+
},
|
| 5096 |
+
{
|
| 5097 |
+
"epoch": 3.01,
|
| 5098 |
+
"eval_loss": 0.7583074569702148,
|
| 5099 |
+
"eval_runtime": 80.3527,
|
| 5100 |
+
"eval_samples_per_second": 12.445,
|
| 5101 |
+
"eval_steps_per_second": 6.223,
|
| 5102 |
+
"step": 5400
|
| 5103 |
+
},
|
| 5104 |
+
{
|
| 5105 |
+
"epoch": 3.01,
|
| 5106 |
+
"mmlu_eval_accuracy": 0.49393296384365404,
|
| 5107 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
|
| 5108 |
+
"mmlu_eval_accuracy_anatomy": 0.7142857142857143,
|
| 5109 |
+
"mmlu_eval_accuracy_astronomy": 0.5,
|
| 5110 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 5111 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
|
| 5112 |
+
"mmlu_eval_accuracy_college_biology": 0.4375,
|
| 5113 |
+
"mmlu_eval_accuracy_college_chemistry": 0.125,
|
| 5114 |
+
"mmlu_eval_accuracy_college_computer_science": 0.18181818181818182,
|
| 5115 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 5116 |
+
"mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
|
| 5117 |
+
"mmlu_eval_accuracy_college_physics": 0.36363636363636365,
|
| 5118 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 5119 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
|
| 5120 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 5121 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.25,
|
| 5122 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
|
| 5123 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 5124 |
+
"mmlu_eval_accuracy_global_facts": 0.5,
|
| 5125 |
+
"mmlu_eval_accuracy_high_school_biology": 0.375,
|
| 5126 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
|
| 5127 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
|
| 5128 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 5129 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
|
| 5130 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
|
| 5131 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4418604651162791,
|
| 5132 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
|
| 5133 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
|
| 5134 |
+
"mmlu_eval_accuracy_high_school_physics": 0.47058823529411764,
|
| 5135 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.85,
|
| 5136 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
|
| 5137 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.5909090909090909,
|
| 5138 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
|
| 5139 |
+
"mmlu_eval_accuracy_human_aging": 0.6956521739130435,
|
| 5140 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 5141 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 5142 |
+
"mmlu_eval_accuracy_jurisprudence": 0.2727272727272727,
|
| 5143 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
|
| 5144 |
+
"mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
|
| 5145 |
+
"mmlu_eval_accuracy_management": 0.5454545454545454,
|
| 5146 |
+
"mmlu_eval_accuracy_marketing": 0.72,
|
| 5147 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 5148 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
|
| 5149 |
+
"mmlu_eval_accuracy_moral_disputes": 0.42105263157894735,
|
| 5150 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.29,
|
| 5151 |
+
"mmlu_eval_accuracy_nutrition": 0.6666666666666666,
|
| 5152 |
+
"mmlu_eval_accuracy_philosophy": 0.5,
|
| 5153 |
+
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
|
| 5154 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
|
| 5155 |
+
"mmlu_eval_accuracy_professional_law": 0.3058823529411765,
|
| 5156 |
+
"mmlu_eval_accuracy_professional_medicine": 0.6451612903225806,
|
| 5157 |
+
"mmlu_eval_accuracy_professional_psychology": 0.463768115942029,
|
| 5158 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 5159 |
+
"mmlu_eval_accuracy_security_studies": 0.4444444444444444,
|
| 5160 |
+
"mmlu_eval_accuracy_sociology": 0.7272727272727273,
|
| 5161 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 5162 |
+
"mmlu_eval_accuracy_virology": 0.5,
|
| 5163 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 5164 |
+
"mmlu_loss": 1.5723119864809296,
|
| 5165 |
+
"step": 5400
|
| 5166 |
+
},
|
| 5167 |
+
{
|
| 5168 |
+
"epoch": 3.01,
|
| 5169 |
+
"learning_rate": 0.0002,
|
| 5170 |
+
"loss": 0.4638,
|
| 5171 |
+
"step": 5410
|
| 5172 |
+
},
|
| 5173 |
+
{
|
| 5174 |
+
"epoch": 3.02,
|
| 5175 |
+
"learning_rate": 0.0002,
|
| 5176 |
+
"loss": 0.5236,
|
| 5177 |
+
"step": 5420
|
| 5178 |
+
},
|
| 5179 |
+
{
|
| 5180 |
+
"epoch": 3.02,
|
| 5181 |
+
"learning_rate": 0.0002,
|
| 5182 |
+
"loss": 0.4499,
|
| 5183 |
+
"step": 5430
|
| 5184 |
+
},
|
| 5185 |
+
{
|
| 5186 |
+
"epoch": 3.03,
|
| 5187 |
+
"learning_rate": 0.0002,
|
| 5188 |
+
"loss": 0.4545,
|
| 5189 |
+
"step": 5440
|
| 5190 |
+
},
|
| 5191 |
+
{
|
| 5192 |
+
"epoch": 3.03,
|
| 5193 |
+
"learning_rate": 0.0002,
|
| 5194 |
+
"loss": 0.5577,
|
| 5195 |
+
"step": 5450
|
| 5196 |
+
},
|
| 5197 |
+
{
|
| 5198 |
+
"epoch": 3.04,
|
| 5199 |
+
"learning_rate": 0.0002,
|
| 5200 |
+
"loss": 0.4587,
|
| 5201 |
+
"step": 5460
|
| 5202 |
+
},
|
| 5203 |
+
{
|
| 5204 |
+
"epoch": 3.04,
|
| 5205 |
+
"learning_rate": 0.0002,
|
| 5206 |
+
"loss": 0.4858,
|
| 5207 |
+
"step": 5470
|
| 5208 |
+
},
|
| 5209 |
+
{
|
| 5210 |
+
"epoch": 3.05,
|
| 5211 |
+
"learning_rate": 0.0002,
|
| 5212 |
+
"loss": 0.5282,
|
| 5213 |
+
"step": 5480
|
| 5214 |
+
},
|
| 5215 |
+
{
|
| 5216 |
+
"epoch": 3.06,
|
| 5217 |
+
"learning_rate": 0.0002,
|
| 5218 |
+
"loss": 0.4659,
|
| 5219 |
+
"step": 5490
|
| 5220 |
+
},
|
| 5221 |
+
{
|
| 5222 |
+
"epoch": 3.06,
|
| 5223 |
+
"learning_rate": 0.0002,
|
| 5224 |
+
"loss": 0.4843,
|
| 5225 |
+
"step": 5500
|
| 5226 |
+
},
|
| 5227 |
+
{
|
| 5228 |
+
"epoch": 3.07,
|
| 5229 |
+
"learning_rate": 0.0002,
|
| 5230 |
+
"loss": 0.4933,
|
| 5231 |
+
"step": 5510
|
| 5232 |
+
},
|
| 5233 |
+
{
|
| 5234 |
+
"epoch": 3.07,
|
| 5235 |
+
"learning_rate": 0.0002,
|
| 5236 |
+
"loss": 0.5252,
|
| 5237 |
+
"step": 5520
|
| 5238 |
+
},
|
| 5239 |
+
{
|
| 5240 |
+
"epoch": 3.08,
|
| 5241 |
+
"learning_rate": 0.0002,
|
| 5242 |
+
"loss": 0.4685,
|
| 5243 |
+
"step": 5530
|
| 5244 |
+
},
|
| 5245 |
+
{
|
| 5246 |
+
"epoch": 3.08,
|
| 5247 |
+
"learning_rate": 0.0002,
|
| 5248 |
+
"loss": 0.5182,
|
| 5249 |
+
"step": 5540
|
| 5250 |
+
},
|
| 5251 |
+
{
|
| 5252 |
+
"epoch": 3.09,
|
| 5253 |
+
"learning_rate": 0.0002,
|
| 5254 |
+
"loss": 0.4584,
|
| 5255 |
+
"step": 5550
|
| 5256 |
+
},
|
| 5257 |
+
{
|
| 5258 |
+
"epoch": 3.09,
|
| 5259 |
+
"learning_rate": 0.0002,
|
| 5260 |
+
"loss": 0.5359,
|
| 5261 |
+
"step": 5560
|
| 5262 |
+
},
|
| 5263 |
+
{
|
| 5264 |
+
"epoch": 3.1,
|
| 5265 |
+
"learning_rate": 0.0002,
|
| 5266 |
+
"loss": 0.4764,
|
| 5267 |
+
"step": 5570
|
| 5268 |
+
},
|
| 5269 |
+
{
|
| 5270 |
+
"epoch": 3.11,
|
| 5271 |
+
"learning_rate": 0.0002,
|
| 5272 |
+
"loss": 0.4843,
|
| 5273 |
+
"step": 5580
|
| 5274 |
+
},
|
| 5275 |
+
{
|
| 5276 |
+
"epoch": 3.11,
|
| 5277 |
+
"learning_rate": 0.0002,
|
| 5278 |
+
"loss": 0.5357,
|
| 5279 |
+
"step": 5590
|
| 5280 |
+
},
|
| 5281 |
+
{
|
| 5282 |
+
"epoch": 3.12,
|
| 5283 |
+
"learning_rate": 0.0002,
|
| 5284 |
+
"loss": 0.5134,
|
| 5285 |
+
"step": 5600
|
| 5286 |
+
},
|
| 5287 |
+
{
|
| 5288 |
+
"epoch": 3.12,
|
| 5289 |
+
"eval_loss": 0.7698976397514343,
|
| 5290 |
+
"eval_runtime": 81.0335,
|
| 5291 |
+
"eval_samples_per_second": 12.341,
|
| 5292 |
+
"eval_steps_per_second": 6.17,
|
| 5293 |
+
"step": 5600
|
| 5294 |
+
},
|
| 5295 |
+
{
|
| 5296 |
+
"epoch": 3.12,
|
| 5297 |
+
"mmlu_eval_accuracy": 0.4999734612828784,
|
| 5298 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
|
| 5299 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 5300 |
+
"mmlu_eval_accuracy_astronomy": 0.4375,
|
| 5301 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 5302 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
|
| 5303 |
+
"mmlu_eval_accuracy_college_biology": 0.5,
|
| 5304 |
+
"mmlu_eval_accuracy_college_chemistry": 0.0,
|
| 5305 |
+
"mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
|
| 5306 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 5307 |
+
"mmlu_eval_accuracy_college_medicine": 0.36363636363636365,
|
| 5308 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 5309 |
+
"mmlu_eval_accuracy_computer_security": 0.36363636363636365,
|
| 5310 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 5311 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 5312 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.3125,
|
| 5313 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
|
| 5314 |
+
"mmlu_eval_accuracy_formal_logic": 0.07142857142857142,
|
| 5315 |
+
"mmlu_eval_accuracy_global_facts": 0.7,
|
| 5316 |
+
"mmlu_eval_accuracy_high_school_biology": 0.4375,
|
| 5317 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
|
| 5318 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
|
| 5319 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
|
| 5320 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
|
| 5321 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
|
| 5322 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.46511627906976744,
|
| 5323 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
|
| 5324 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
|
| 5325 |
+
"mmlu_eval_accuracy_high_school_physics": 0.4117647058823529,
|
| 5326 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.85,
|
| 5327 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
|
| 5328 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.5454545454545454,
|
| 5329 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
|
| 5330 |
+
"mmlu_eval_accuracy_human_aging": 0.6521739130434783,
|
| 5331 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 5332 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 5333 |
+
"mmlu_eval_accuracy_jurisprudence": 0.2727272727272727,
|
| 5334 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
|
| 5335 |
+
"mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
|
| 5336 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 5337 |
+
"mmlu_eval_accuracy_marketing": 0.8,
|
| 5338 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 5339 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
|
| 5340 |
+
"mmlu_eval_accuracy_moral_disputes": 0.5,
|
| 5341 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.26,
|
| 5342 |
+
"mmlu_eval_accuracy_nutrition": 0.7272727272727273,
|
| 5343 |
+
"mmlu_eval_accuracy_philosophy": 0.5294117647058824,
|
| 5344 |
+
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
|
| 5345 |
+
"mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
|
| 5346 |
+
"mmlu_eval_accuracy_professional_law": 0.29411764705882354,
|
| 5347 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
|
| 5348 |
+
"mmlu_eval_accuracy_professional_psychology": 0.43478260869565216,
|
| 5349 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 5350 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 5351 |
+
"mmlu_eval_accuracy_sociology": 0.8181818181818182,
|
| 5352 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
|
| 5353 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 5354 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 5355 |
+
"mmlu_loss": 1.678168613667899,
|
| 5356 |
+
"step": 5600
|
| 5357 |
+
},
|
| 5358 |
+
{
|
| 5359 |
+
"epoch": 3.12,
|
| 5360 |
+
"learning_rate": 0.0002,
|
| 5361 |
+
"loss": 0.4674,
|
| 5362 |
+
"step": 5610
|
| 5363 |
+
},
|
| 5364 |
+
{
|
| 5365 |
+
"epoch": 3.13,
|
| 5366 |
+
"learning_rate": 0.0002,
|
| 5367 |
+
"loss": 0.5213,
|
| 5368 |
+
"step": 5620
|
| 5369 |
+
},
|
| 5370 |
+
{
|
| 5371 |
+
"epoch": 3.13,
|
| 5372 |
+
"learning_rate": 0.0002,
|
| 5373 |
+
"loss": 0.5247,
|
| 5374 |
+
"step": 5630
|
| 5375 |
+
},
|
| 5376 |
+
{
|
| 5377 |
+
"epoch": 3.14,
|
| 5378 |
+
"learning_rate": 0.0002,
|
| 5379 |
+
"loss": 0.4982,
|
| 5380 |
+
"step": 5640
|
| 5381 |
+
},
|
| 5382 |
+
{
|
| 5383 |
+
"epoch": 3.14,
|
| 5384 |
+
"learning_rate": 0.0002,
|
| 5385 |
+
"loss": 0.5065,
|
| 5386 |
+
"step": 5650
|
| 5387 |
+
},
|
| 5388 |
+
{
|
| 5389 |
+
"epoch": 3.15,
|
| 5390 |
+
"learning_rate": 0.0002,
|
| 5391 |
+
"loss": 0.5408,
|
| 5392 |
+
"step": 5660
|
| 5393 |
+
},
|
| 5394 |
+
{
|
| 5395 |
+
"epoch": 3.16,
|
| 5396 |
+
"learning_rate": 0.0002,
|
| 5397 |
+
"loss": 0.4885,
|
| 5398 |
+
"step": 5670
|
| 5399 |
+
},
|
| 5400 |
+
{
|
| 5401 |
+
"epoch": 3.16,
|
| 5402 |
+
"learning_rate": 0.0002,
|
| 5403 |
+
"loss": 0.5337,
|
| 5404 |
+
"step": 5680
|
| 5405 |
+
},
|
| 5406 |
+
{
|
| 5407 |
+
"epoch": 3.17,
|
| 5408 |
+
"learning_rate": 0.0002,
|
| 5409 |
+
"loss": 0.4754,
|
| 5410 |
+
"step": 5690
|
| 5411 |
+
},
|
| 5412 |
+
{
|
| 5413 |
+
"epoch": 3.17,
|
| 5414 |
+
"learning_rate": 0.0002,
|
| 5415 |
+
"loss": 0.5275,
|
| 5416 |
+
"step": 5700
|
| 5417 |
+
},
|
| 5418 |
+
{
|
| 5419 |
+
"epoch": 3.18,
|
| 5420 |
+
"learning_rate": 0.0002,
|
| 5421 |
+
"loss": 0.4626,
|
| 5422 |
+
"step": 5710
|
| 5423 |
+
},
|
| 5424 |
+
{
|
| 5425 |
+
"epoch": 3.18,
|
| 5426 |
+
"learning_rate": 0.0002,
|
| 5427 |
+
"loss": 0.4585,
|
| 5428 |
+
"step": 5720
|
| 5429 |
+
},
|
| 5430 |
+
{
|
| 5431 |
+
"epoch": 3.19,
|
| 5432 |
+
"learning_rate": 0.0002,
|
| 5433 |
+
"loss": 0.439,
|
| 5434 |
+
"step": 5730
|
| 5435 |
+
},
|
| 5436 |
+
{
|
| 5437 |
+
"epoch": 3.19,
|
| 5438 |
+
"learning_rate": 0.0002,
|
| 5439 |
+
"loss": 0.5295,
|
| 5440 |
+
"step": 5740
|
| 5441 |
+
},
|
| 5442 |
+
{
|
| 5443 |
+
"epoch": 3.2,
|
| 5444 |
+
"learning_rate": 0.0002,
|
| 5445 |
+
"loss": 0.4673,
|
| 5446 |
+
"step": 5750
|
| 5447 |
+
},
|
| 5448 |
+
{
|
| 5449 |
+
"epoch": 3.21,
|
| 5450 |
+
"learning_rate": 0.0002,
|
| 5451 |
+
"loss": 0.4648,
|
| 5452 |
+
"step": 5760
|
| 5453 |
+
},
|
| 5454 |
+
{
|
| 5455 |
+
"epoch": 3.21,
|
| 5456 |
+
"learning_rate": 0.0002,
|
| 5457 |
+
"loss": 0.5361,
|
| 5458 |
+
"step": 5770
|
| 5459 |
+
},
|
| 5460 |
+
{
|
| 5461 |
+
"epoch": 3.22,
|
| 5462 |
+
"learning_rate": 0.0002,
|
| 5463 |
+
"loss": 0.4478,
|
| 5464 |
+
"step": 5780
|
| 5465 |
+
},
|
| 5466 |
+
{
|
| 5467 |
+
"epoch": 3.22,
|
| 5468 |
+
"learning_rate": 0.0002,
|
| 5469 |
+
"loss": 0.4774,
|
| 5470 |
+
"step": 5790
|
| 5471 |
+
},
|
| 5472 |
+
{
|
| 5473 |
+
"epoch": 3.23,
|
| 5474 |
+
"learning_rate": 0.0002,
|
| 5475 |
+
"loss": 0.4614,
|
| 5476 |
+
"step": 5800
|
| 5477 |
+
},
|
| 5478 |
+
{
|
| 5479 |
+
"epoch": 3.23,
|
| 5480 |
+
"eval_loss": 0.7794032096862793,
|
| 5481 |
+
"eval_runtime": 80.9662,
|
| 5482 |
+
"eval_samples_per_second": 12.351,
|
| 5483 |
+
"eval_steps_per_second": 6.175,
|
| 5484 |
+
"step": 5800
|
| 5485 |
+
},
|
| 5486 |
+
{
|
| 5487 |
+
"epoch": 3.23,
|
| 5488 |
+
"mmlu_eval_accuracy": 0.48090560036810726,
|
| 5489 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
|
| 5490 |
+
"mmlu_eval_accuracy_anatomy": 0.5,
|
| 5491 |
+
"mmlu_eval_accuracy_astronomy": 0.4375,
|
| 5492 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 5493 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.4482758620689655,
|
| 5494 |
+
"mmlu_eval_accuracy_college_biology": 0.5,
|
| 5495 |
+
"mmlu_eval_accuracy_college_chemistry": 0.0,
|
| 5496 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 5497 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 5498 |
+
"mmlu_eval_accuracy_college_medicine": 0.3181818181818182,
|
| 5499 |
+
"mmlu_eval_accuracy_college_physics": 0.2727272727272727,
|
| 5500 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 5501 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
|
| 5502 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 5503 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.3125,
|
| 5504 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 5505 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 5506 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 5507 |
+
"mmlu_eval_accuracy_high_school_biology": 0.4375,
|
| 5508 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
|
| 5509 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
|
| 5510 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 5511 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
|
| 5512 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
|
| 5513 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
|
| 5514 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.13793103448275862,
|
| 5515 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
|
| 5516 |
+
"mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
|
| 5517 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
|
| 5518 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
|
| 5519 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.5909090909090909,
|
| 5520 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
|
| 5521 |
+
"mmlu_eval_accuracy_human_aging": 0.6521739130434783,
|
| 5522 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 5523 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 5524 |
+
"mmlu_eval_accuracy_jurisprudence": 0.2727272727272727,
|
| 5525 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
|
| 5526 |
+
"mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
|
| 5527 |
+
"mmlu_eval_accuracy_management": 0.5454545454545454,
|
| 5528 |
+
"mmlu_eval_accuracy_marketing": 0.84,
|
| 5529 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 5530 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
|
| 5531 |
+
"mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
|
| 5532 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.31,
|
| 5533 |
+
"mmlu_eval_accuracy_nutrition": 0.6666666666666666,
|
| 5534 |
+
"mmlu_eval_accuracy_philosophy": 0.5294117647058824,
|
| 5535 |
+
"mmlu_eval_accuracy_prehistory": 0.42857142857142855,
|
| 5536 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
|
| 5537 |
+
"mmlu_eval_accuracy_professional_law": 0.3235294117647059,
|
| 5538 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
|
| 5539 |
+
"mmlu_eval_accuracy_professional_psychology": 0.463768115942029,
|
| 5540 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 5541 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 5542 |
+
"mmlu_eval_accuracy_sociology": 0.7272727272727273,
|
| 5543 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
|
| 5544 |
+
"mmlu_eval_accuracy_virology": 0.5,
|
| 5545 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 5546 |
+
"mmlu_loss": 1.6652605916439087,
|
| 5547 |
+
"step": 5800
|
| 5548 |
+
},
|
| 5549 |
+
{
|
| 5550 |
+
"epoch": 3.23,
|
| 5551 |
+
"learning_rate": 0.0002,
|
| 5552 |
+
"loss": 0.4767,
|
| 5553 |
+
"step": 5810
|
| 5554 |
+
},
|
| 5555 |
+
{
|
| 5556 |
+
"epoch": 3.24,
|
| 5557 |
+
"learning_rate": 0.0002,
|
| 5558 |
+
"loss": 0.4987,
|
| 5559 |
+
"step": 5820
|
| 5560 |
+
},
|
| 5561 |
+
{
|
| 5562 |
+
"epoch": 3.24,
|
| 5563 |
+
"learning_rate": 0.0002,
|
| 5564 |
+
"loss": 0.505,
|
| 5565 |
+
"step": 5830
|
| 5566 |
+
},
|
| 5567 |
+
{
|
| 5568 |
+
"epoch": 3.25,
|
| 5569 |
+
"learning_rate": 0.0002,
|
| 5570 |
+
"loss": 0.5065,
|
| 5571 |
+
"step": 5840
|
| 5572 |
+
},
|
| 5573 |
+
{
|
| 5574 |
+
"epoch": 3.26,
|
| 5575 |
+
"learning_rate": 0.0002,
|
| 5576 |
+
"loss": 0.5165,
|
| 5577 |
+
"step": 5850
|
| 5578 |
+
},
|
| 5579 |
+
{
|
| 5580 |
+
"epoch": 3.26,
|
| 5581 |
+
"learning_rate": 0.0002,
|
| 5582 |
+
"loss": 0.5305,
|
| 5583 |
+
"step": 5860
|
| 5584 |
+
},
|
| 5585 |
+
{
|
| 5586 |
+
"epoch": 3.27,
|
| 5587 |
+
"learning_rate": 0.0002,
|
| 5588 |
+
"loss": 0.4807,
|
| 5589 |
+
"step": 5870
|
| 5590 |
+
},
|
| 5591 |
+
{
|
| 5592 |
+
"epoch": 3.27,
|
| 5593 |
+
"learning_rate": 0.0002,
|
| 5594 |
+
"loss": 0.5399,
|
| 5595 |
+
"step": 5880
|
| 5596 |
+
},
|
| 5597 |
+
{
|
| 5598 |
+
"epoch": 3.28,
|
| 5599 |
+
"learning_rate": 0.0002,
|
| 5600 |
+
"loss": 0.5041,
|
| 5601 |
+
"step": 5890
|
| 5602 |
+
},
|
| 5603 |
+
{
|
| 5604 |
+
"epoch": 3.28,
|
| 5605 |
+
"learning_rate": 0.0002,
|
| 5606 |
+
"loss": 0.4844,
|
| 5607 |
+
"step": 5900
|
| 5608 |
+
},
|
| 5609 |
+
{
|
| 5610 |
+
"epoch": 3.29,
|
| 5611 |
+
"learning_rate": 0.0002,
|
| 5612 |
+
"loss": 0.4823,
|
| 5613 |
+
"step": 5910
|
| 5614 |
+
},
|
| 5615 |
+
{
|
| 5616 |
+
"epoch": 3.29,
|
| 5617 |
+
"learning_rate": 0.0002,
|
| 5618 |
+
"loss": 0.5222,
|
| 5619 |
+
"step": 5920
|
| 5620 |
+
},
|
| 5621 |
+
{
|
| 5622 |
+
"epoch": 3.3,
|
| 5623 |
+
"learning_rate": 0.0002,
|
| 5624 |
+
"loss": 0.5115,
|
| 5625 |
+
"step": 5930
|
| 5626 |
+
},
|
| 5627 |
+
{
|
| 5628 |
+
"epoch": 3.31,
|
| 5629 |
+
"learning_rate": 0.0002,
|
| 5630 |
+
"loss": 0.5214,
|
| 5631 |
+
"step": 5940
|
| 5632 |
+
},
|
| 5633 |
+
{
|
| 5634 |
+
"epoch": 3.31,
|
| 5635 |
+
"learning_rate": 0.0002,
|
| 5636 |
+
"loss": 0.4697,
|
| 5637 |
+
"step": 5950
|
| 5638 |
+
},
|
| 5639 |
+
{
|
| 5640 |
+
"epoch": 3.32,
|
| 5641 |
+
"learning_rate": 0.0002,
|
| 5642 |
+
"loss": 0.599,
|
| 5643 |
+
"step": 5960
|
| 5644 |
+
},
|
| 5645 |
+
{
|
| 5646 |
+
"epoch": 3.32,
|
| 5647 |
+
"learning_rate": 0.0002,
|
| 5648 |
+
"loss": 0.5582,
|
| 5649 |
+
"step": 5970
|
| 5650 |
+
},
|
| 5651 |
+
{
|
| 5652 |
+
"epoch": 3.33,
|
| 5653 |
+
"learning_rate": 0.0002,
|
| 5654 |
+
"loss": 0.4939,
|
| 5655 |
+
"step": 5980
|
| 5656 |
+
},
|
| 5657 |
+
{
|
| 5658 |
+
"epoch": 3.33,
|
| 5659 |
+
"learning_rate": 0.0002,
|
| 5660 |
+
"loss": 0.5328,
|
| 5661 |
+
"step": 5990
|
| 5662 |
+
},
|
| 5663 |
+
{
|
| 5664 |
+
"epoch": 3.34,
|
| 5665 |
+
"learning_rate": 0.0002,
|
| 5666 |
+
"loss": 0.58,
|
| 5667 |
+
"step": 6000
|
| 5668 |
+
},
|
| 5669 |
+
{
|
| 5670 |
+
"epoch": 3.34,
|
| 5671 |
+
"eval_loss": 0.7839476466178894,
|
| 5672 |
+
"eval_runtime": 80.9237,
|
| 5673 |
+
"eval_samples_per_second": 12.357,
|
| 5674 |
+
"eval_steps_per_second": 6.179,
|
| 5675 |
+
"step": 6000
|
| 5676 |
+
},
|
| 5677 |
+
{
|
| 5678 |
+
"epoch": 3.34,
|
| 5679 |
+
"mmlu_eval_accuracy": 0.4762839451968676,
|
| 5680 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
|
| 5681 |
+
"mmlu_eval_accuracy_anatomy": 0.5714285714285714,
|
| 5682 |
+
"mmlu_eval_accuracy_astronomy": 0.375,
|
| 5683 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 5684 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.4482758620689655,
|
| 5685 |
+
"mmlu_eval_accuracy_college_biology": 0.4375,
|
| 5686 |
+
"mmlu_eval_accuracy_college_chemistry": 0.0,
|
| 5687 |
+
"mmlu_eval_accuracy_college_computer_science": 0.09090909090909091,
|
| 5688 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 5689 |
+
"mmlu_eval_accuracy_college_medicine": 0.36363636363636365,
|
| 5690 |
+
"mmlu_eval_accuracy_college_physics": 0.2727272727272727,
|
| 5691 |
+
"mmlu_eval_accuracy_computer_security": 0.36363636363636365,
|
| 5692 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 5693 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 5694 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.25,
|
| 5695 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 5696 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 5697 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 5698 |
+
"mmlu_eval_accuracy_high_school_biology": 0.40625,
|
| 5699 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
|
| 5700 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
|
| 5701 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
|
| 5702 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
|
| 5703 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
|
| 5704 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
|
| 5705 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
|
| 5706 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.34615384615384615,
|
| 5707 |
+
"mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
|
| 5708 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
|
| 5709 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
|
| 5710 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.5909090909090909,
|
| 5711 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
|
| 5712 |
+
"mmlu_eval_accuracy_human_aging": 0.5652173913043478,
|
| 5713 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 5714 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 5715 |
+
"mmlu_eval_accuracy_jurisprudence": 0.2727272727272727,
|
| 5716 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
|
| 5717 |
+
"mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
|
| 5718 |
+
"mmlu_eval_accuracy_management": 0.5454545454545454,
|
| 5719 |
+
"mmlu_eval_accuracy_marketing": 0.8,
|
| 5720 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 5721 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
|
| 5722 |
+
"mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
|
| 5723 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.3,
|
| 5724 |
+
"mmlu_eval_accuracy_nutrition": 0.6666666666666666,
|
| 5725 |
+
"mmlu_eval_accuracy_philosophy": 0.5294117647058824,
|
| 5726 |
+
"mmlu_eval_accuracy_prehistory": 0.42857142857142855,
|
| 5727 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
|
| 5728 |
+
"mmlu_eval_accuracy_professional_law": 0.3058823529411765,
|
| 5729 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
|
| 5730 |
+
"mmlu_eval_accuracy_professional_psychology": 0.463768115942029,
|
| 5731 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 5732 |
+
"mmlu_eval_accuracy_security_studies": 0.4444444444444444,
|
| 5733 |
+
"mmlu_eval_accuracy_sociology": 0.8181818181818182,
|
| 5734 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
|
| 5735 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 5736 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 5737 |
+
"mmlu_loss": 1.795020043305255,
|
| 5738 |
+
"step": 6000
|
| 5739 |
+
},
|
| 5740 |
+
{
|
| 5741 |
+
"epoch": 3.34,
|
| 5742 |
+
"learning_rate": 0.0002,
|
| 5743 |
+
"loss": 0.4693,
|
| 5744 |
+
"step": 6010
|
| 5745 |
+
},
|
| 5746 |
+
{
|
| 5747 |
+
"epoch": 3.35,
|
| 5748 |
+
"learning_rate": 0.0002,
|
| 5749 |
+
"loss": 0.4513,
|
| 5750 |
+
"step": 6020
|
| 5751 |
+
},
|
| 5752 |
+
{
|
| 5753 |
+
"epoch": 3.36,
|
| 5754 |
+
"learning_rate": 0.0002,
|
| 5755 |
+
"loss": 0.5278,
|
| 5756 |
+
"step": 6030
|
| 5757 |
+
},
|
| 5758 |
+
{
|
| 5759 |
+
"epoch": 3.36,
|
| 5760 |
+
"learning_rate": 0.0002,
|
| 5761 |
+
"loss": 0.5482,
|
| 5762 |
+
"step": 6040
|
| 5763 |
+
},
|
| 5764 |
+
{
|
| 5765 |
+
"epoch": 3.37,
|
| 5766 |
+
"learning_rate": 0.0002,
|
| 5767 |
+
"loss": 0.4568,
|
| 5768 |
+
"step": 6050
|
| 5769 |
+
},
|
| 5770 |
+
{
|
| 5771 |
+
"epoch": 3.37,
|
| 5772 |
+
"learning_rate": 0.0002,
|
| 5773 |
+
"loss": 0.4826,
|
| 5774 |
+
"step": 6060
|
| 5775 |
+
},
|
| 5776 |
+
{
|
| 5777 |
+
"epoch": 3.38,
|
| 5778 |
+
"learning_rate": 0.0002,
|
| 5779 |
+
"loss": 0.4501,
|
| 5780 |
+
"step": 6070
|
| 5781 |
+
},
|
| 5782 |
+
{
|
| 5783 |
+
"epoch": 3.38,
|
| 5784 |
+
"learning_rate": 0.0002,
|
| 5785 |
+
"loss": 0.4835,
|
| 5786 |
+
"step": 6080
|
| 5787 |
+
},
|
| 5788 |
+
{
|
| 5789 |
+
"epoch": 3.39,
|
| 5790 |
+
"learning_rate": 0.0002,
|
| 5791 |
+
"loss": 0.5297,
|
| 5792 |
+
"step": 6090
|
| 5793 |
+
},
|
| 5794 |
+
{
|
| 5795 |
+
"epoch": 3.39,
|
| 5796 |
+
"learning_rate": 0.0002,
|
| 5797 |
+
"loss": 0.5004,
|
| 5798 |
+
"step": 6100
|
| 5799 |
+
},
|
| 5800 |
+
{
|
| 5801 |
+
"epoch": 3.4,
|
| 5802 |
+
"learning_rate": 0.0002,
|
| 5803 |
+
"loss": 0.5338,
|
| 5804 |
+
"step": 6110
|
| 5805 |
+
},
|
| 5806 |
+
{
|
| 5807 |
+
"epoch": 3.41,
|
| 5808 |
+
"learning_rate": 0.0002,
|
| 5809 |
+
"loss": 0.47,
|
| 5810 |
+
"step": 6120
|
| 5811 |
+
},
|
| 5812 |
+
{
|
| 5813 |
+
"epoch": 3.41,
|
| 5814 |
+
"learning_rate": 0.0002,
|
| 5815 |
+
"loss": 0.5506,
|
| 5816 |
+
"step": 6130
|
| 5817 |
+
},
|
| 5818 |
+
{
|
| 5819 |
+
"epoch": 3.42,
|
| 5820 |
+
"learning_rate": 0.0002,
|
| 5821 |
+
"loss": 0.5345,
|
| 5822 |
+
"step": 6140
|
| 5823 |
+
},
|
| 5824 |
+
{
|
| 5825 |
+
"epoch": 3.42,
|
| 5826 |
+
"learning_rate": 0.0002,
|
| 5827 |
+
"loss": 0.4767,
|
| 5828 |
+
"step": 6150
|
| 5829 |
+
},
|
| 5830 |
+
{
|
| 5831 |
+
"epoch": 3.43,
|
| 5832 |
+
"learning_rate": 0.0002,
|
| 5833 |
+
"loss": 0.5117,
|
| 5834 |
+
"step": 6160
|
| 5835 |
+
},
|
| 5836 |
+
{
|
| 5837 |
+
"epoch": 3.43,
|
| 5838 |
+
"learning_rate": 0.0002,
|
| 5839 |
+
"loss": 0.551,
|
| 5840 |
+
"step": 6170
|
| 5841 |
+
},
|
| 5842 |
+
{
|
| 5843 |
+
"epoch": 3.44,
|
| 5844 |
+
"learning_rate": 0.0002,
|
| 5845 |
+
"loss": 0.5547,
|
| 5846 |
+
"step": 6180
|
| 5847 |
+
},
|
| 5848 |
+
{
|
| 5849 |
+
"epoch": 3.44,
|
| 5850 |
+
"learning_rate": 0.0002,
|
| 5851 |
+
"loss": 0.4606,
|
| 5852 |
+
"step": 6190
|
| 5853 |
+
},
|
| 5854 |
+
{
|
| 5855 |
+
"epoch": 3.45,
|
| 5856 |
+
"learning_rate": 0.0002,
|
| 5857 |
+
"loss": 0.5563,
|
| 5858 |
+
"step": 6200
|
| 5859 |
+
},
|
| 5860 |
+
{
|
| 5861 |
+
"epoch": 3.45,
|
| 5862 |
+
"eval_loss": 0.764141857624054,
|
| 5863 |
+
"eval_runtime": 80.9038,
|
| 5864 |
+
"eval_samples_per_second": 12.36,
|
| 5865 |
+
"eval_steps_per_second": 6.18,
|
| 5866 |
+
"step": 6200
|
| 5867 |
+
},
|
| 5868 |
+
{
|
| 5869 |
+
"epoch": 3.45,
|
| 5870 |
+
"mmlu_eval_accuracy": 0.47738388369750223,
|
| 5871 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
|
| 5872 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 5873 |
+
"mmlu_eval_accuracy_astronomy": 0.4375,
|
| 5874 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 5875 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.4482758620689655,
|
| 5876 |
+
"mmlu_eval_accuracy_college_biology": 0.5625,
|
| 5877 |
+
"mmlu_eval_accuracy_college_chemistry": 0.0,
|
| 5878 |
+
"mmlu_eval_accuracy_college_computer_science": 0.18181818181818182,
|
| 5879 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 5880 |
+
"mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
|
| 5881 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 5882 |
+
"mmlu_eval_accuracy_computer_security": 0.36363636363636365,
|
| 5883 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 5884 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 5885 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.1875,
|
| 5886 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 5887 |
+
"mmlu_eval_accuracy_formal_logic": 0.07142857142857142,
|
| 5888 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 5889 |
+
"mmlu_eval_accuracy_high_school_biology": 0.40625,
|
| 5890 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
|
| 5891 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
|
| 5892 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
|
| 5893 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
|
| 5894 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
|
| 5895 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.46511627906976744,
|
| 5896 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
|
| 5897 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
|
| 5898 |
+
"mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
|
| 5899 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
|
| 5900 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
|
| 5901 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.5909090909090909,
|
| 5902 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
|
| 5903 |
+
"mmlu_eval_accuracy_human_aging": 0.6521739130434783,
|
| 5904 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 5905 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 5906 |
+
"mmlu_eval_accuracy_jurisprudence": 0.2727272727272727,
|
| 5907 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
|
| 5908 |
+
"mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
|
| 5909 |
+
"mmlu_eval_accuracy_management": 0.5454545454545454,
|
| 5910 |
+
"mmlu_eval_accuracy_marketing": 0.88,
|
| 5911 |
+
"mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
|
| 5912 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
|
| 5913 |
+
"mmlu_eval_accuracy_moral_disputes": 0.5,
|
| 5914 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.35,
|
| 5915 |
+
"mmlu_eval_accuracy_nutrition": 0.6666666666666666,
|
| 5916 |
+
"mmlu_eval_accuracy_philosophy": 0.5294117647058824,
|
| 5917 |
+
"mmlu_eval_accuracy_prehistory": 0.45714285714285713,
|
| 5918 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
|
| 5919 |
+
"mmlu_eval_accuracy_professional_law": 0.3352941176470588,
|
| 5920 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
|
| 5921 |
+
"mmlu_eval_accuracy_professional_psychology": 0.4492753623188406,
|
| 5922 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 5923 |
+
"mmlu_eval_accuracy_security_studies": 0.4444444444444444,
|
| 5924 |
+
"mmlu_eval_accuracy_sociology": 0.7727272727272727,
|
| 5925 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
|
| 5926 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 5927 |
+
"mmlu_eval_accuracy_world_religions": 0.631578947368421,
|
| 5928 |
+
"mmlu_loss": 1.9604056896179836,
|
| 5929 |
+
"step": 6200
|
| 5930 |
+
},
|
| 5931 |
+
{
|
| 5932 |
+
"epoch": 3.46,
|
| 5933 |
+
"learning_rate": 0.0002,
|
| 5934 |
+
"loss": 0.5133,
|
| 5935 |
+
"step": 6210
|
| 5936 |
+
},
|
| 5937 |
+
{
|
| 5938 |
+
"epoch": 3.46,
|
| 5939 |
+
"learning_rate": 0.0002,
|
| 5940 |
+
"loss": 0.4482,
|
| 5941 |
+
"step": 6220
|
| 5942 |
+
},
|
| 5943 |
+
{
|
| 5944 |
+
"epoch": 3.47,
|
| 5945 |
+
"learning_rate": 0.0002,
|
| 5946 |
+
"loss": 0.4692,
|
| 5947 |
+
"step": 6230
|
| 5948 |
+
},
|
| 5949 |
+
{
|
| 5950 |
+
"epoch": 3.47,
|
| 5951 |
+
"learning_rate": 0.0002,
|
| 5952 |
+
"loss": 0.5156,
|
| 5953 |
+
"step": 6240
|
| 5954 |
+
},
|
| 5955 |
+
{
|
| 5956 |
+
"epoch": 3.48,
|
| 5957 |
+
"learning_rate": 0.0002,
|
| 5958 |
+
"loss": 0.4411,
|
| 5959 |
+
"step": 6250
|
| 5960 |
+
},
|
| 5961 |
+
{
|
| 5962 |
+
"epoch": 3.48,
|
| 5963 |
+
"learning_rate": 0.0002,
|
| 5964 |
+
"loss": 0.5514,
|
| 5965 |
+
"step": 6260
|
| 5966 |
+
},
|
| 5967 |
+
{
|
| 5968 |
+
"epoch": 3.49,
|
| 5969 |
+
"learning_rate": 0.0002,
|
| 5970 |
+
"loss": 0.5847,
|
| 5971 |
+
"step": 6270
|
| 5972 |
+
},
|
| 5973 |
+
{
|
| 5974 |
+
"epoch": 3.49,
|
| 5975 |
+
"learning_rate": 0.0002,
|
| 5976 |
+
"loss": 0.4943,
|
| 5977 |
+
"step": 6280
|
| 5978 |
+
},
|
| 5979 |
+
{
|
| 5980 |
+
"epoch": 3.5,
|
| 5981 |
+
"learning_rate": 0.0002,
|
| 5982 |
+
"loss": 0.5319,
|
| 5983 |
+
"step": 6290
|
| 5984 |
+
},
|
| 5985 |
+
{
|
| 5986 |
+
"epoch": 3.51,
|
| 5987 |
+
"learning_rate": 0.0002,
|
| 5988 |
+
"loss": 0.5023,
|
| 5989 |
+
"step": 6300
|
| 5990 |
+
},
|
| 5991 |
+
{
|
| 5992 |
+
"epoch": 3.51,
|
| 5993 |
+
"learning_rate": 0.0002,
|
| 5994 |
+
"loss": 0.5005,
|
| 5995 |
+
"step": 6310
|
| 5996 |
+
},
|
| 5997 |
+
{
|
| 5998 |
+
"epoch": 3.52,
|
| 5999 |
+
"learning_rate": 0.0002,
|
| 6000 |
+
"loss": 0.4763,
|
| 6001 |
+
"step": 6320
|
| 6002 |
+
},
|
| 6003 |
+
{
|
| 6004 |
+
"epoch": 3.52,
|
| 6005 |
+
"learning_rate": 0.0002,
|
| 6006 |
+
"loss": 0.4631,
|
| 6007 |
+
"step": 6330
|
| 6008 |
+
},
|
| 6009 |
+
{
|
| 6010 |
+
"epoch": 3.53,
|
| 6011 |
+
"learning_rate": 0.0002,
|
| 6012 |
+
"loss": 0.5383,
|
| 6013 |
+
"step": 6340
|
| 6014 |
+
},
|
| 6015 |
+
{
|
| 6016 |
+
"epoch": 3.53,
|
| 6017 |
+
"learning_rate": 0.0002,
|
| 6018 |
+
"loss": 0.48,
|
| 6019 |
+
"step": 6350
|
| 6020 |
+
},
|
| 6021 |
+
{
|
| 6022 |
+
"epoch": 3.54,
|
| 6023 |
+
"learning_rate": 0.0002,
|
| 6024 |
+
"loss": 0.4932,
|
| 6025 |
+
"step": 6360
|
| 6026 |
+
},
|
| 6027 |
+
{
|
| 6028 |
+
"epoch": 3.54,
|
| 6029 |
+
"learning_rate": 0.0002,
|
| 6030 |
+
"loss": 0.5599,
|
| 6031 |
+
"step": 6370
|
| 6032 |
+
},
|
| 6033 |
+
{
|
| 6034 |
+
"epoch": 3.55,
|
| 6035 |
+
"learning_rate": 0.0002,
|
| 6036 |
+
"loss": 0.4724,
|
| 6037 |
+
"step": 6380
|
| 6038 |
+
},
|
| 6039 |
+
{
|
| 6040 |
+
"epoch": 3.56,
|
| 6041 |
+
"learning_rate": 0.0002,
|
| 6042 |
+
"loss": 0.4755,
|
| 6043 |
+
"step": 6390
|
| 6044 |
+
},
|
| 6045 |
+
{
|
| 6046 |
+
"epoch": 3.56,
|
| 6047 |
+
"learning_rate": 0.0002,
|
| 6048 |
+
"loss": 0.4781,
|
| 6049 |
+
"step": 6400
|
| 6050 |
+
},
|
| 6051 |
+
{
|
| 6052 |
+
"epoch": 3.56,
|
| 6053 |
+
"eval_loss": 0.7837917804718018,
|
| 6054 |
+
"eval_runtime": 81.0791,
|
| 6055 |
+
"eval_samples_per_second": 12.334,
|
| 6056 |
+
"eval_steps_per_second": 6.167,
|
| 6057 |
+
"step": 6400
|
| 6058 |
+
},
|
| 6059 |
+
{
|
| 6060 |
+
"epoch": 3.56,
|
| 6061 |
+
"mmlu_eval_accuracy": 0.4814930340985633,
|
| 6062 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
|
| 6063 |
+
"mmlu_eval_accuracy_anatomy": 0.5714285714285714,
|
| 6064 |
+
"mmlu_eval_accuracy_astronomy": 0.375,
|
| 6065 |
+
"mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
|
| 6066 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.4482758620689655,
|
| 6067 |
+
"mmlu_eval_accuracy_college_biology": 0.4375,
|
| 6068 |
+
"mmlu_eval_accuracy_college_chemistry": 0.0,
|
| 6069 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 6070 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 6071 |
+
"mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
|
| 6072 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 6073 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 6074 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
|
| 6075 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 6076 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.1875,
|
| 6077 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
|
| 6078 |
+
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
|
| 6079 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 6080 |
+
"mmlu_eval_accuracy_high_school_biology": 0.4375,
|
| 6081 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
|
| 6082 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 6083 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 6084 |
+
"mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
|
| 6085 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
|
| 6086 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4883720930232558,
|
| 6087 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
|
| 6088 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
|
| 6089 |
+
"mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
|
| 6090 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8,
|
| 6091 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
|
| 6092 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.5454545454545454,
|
| 6093 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
|
| 6094 |
+
"mmlu_eval_accuracy_human_aging": 0.6521739130434783,
|
| 6095 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 6096 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 6097 |
+
"mmlu_eval_accuracy_jurisprudence": 0.2727272727272727,
|
| 6098 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
|
| 6099 |
+
"mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
|
| 6100 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 6101 |
+
"mmlu_eval_accuracy_marketing": 0.88,
|
| 6102 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 6103 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
|
| 6104 |
+
"mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
|
| 6105 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.24,
|
| 6106 |
+
"mmlu_eval_accuracy_nutrition": 0.6060606060606061,
|
| 6107 |
+
"mmlu_eval_accuracy_philosophy": 0.5,
|
| 6108 |
+
"mmlu_eval_accuracy_prehistory": 0.45714285714285713,
|
| 6109 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
|
| 6110 |
+
"mmlu_eval_accuracy_professional_law": 0.3176470588235294,
|
| 6111 |
+
"mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
|
| 6112 |
+
"mmlu_eval_accuracy_professional_psychology": 0.4492753623188406,
|
| 6113 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 6114 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 6115 |
+
"mmlu_eval_accuracy_sociology": 0.7727272727272727,
|
| 6116 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
|
| 6117 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 6118 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 6119 |
+
"mmlu_loss": 1.88181437108442,
|
| 6120 |
+
"step": 6400
|
| 6121 |
+
},
|
| 6122 |
+
{
|
| 6123 |
+
"epoch": 3.57,
|
| 6124 |
+
"learning_rate": 0.0002,
|
| 6125 |
+
"loss": 0.4542,
|
| 6126 |
+
"step": 6410
|
| 6127 |
+
},
|
| 6128 |
+
{
|
| 6129 |
+
"epoch": 3.57,
|
| 6130 |
+
"learning_rate": 0.0002,
|
| 6131 |
+
"loss": 0.5077,
|
| 6132 |
+
"step": 6420
|
| 6133 |
+
},
|
| 6134 |
+
{
|
| 6135 |
+
"epoch": 3.58,
|
| 6136 |
+
"learning_rate": 0.0002,
|
| 6137 |
+
"loss": 0.4698,
|
| 6138 |
+
"step": 6430
|
| 6139 |
+
},
|
| 6140 |
+
{
|
| 6141 |
+
"epoch": 3.58,
|
| 6142 |
+
"learning_rate": 0.0002,
|
| 6143 |
+
"loss": 0.4769,
|
| 6144 |
+
"step": 6440
|
| 6145 |
+
},
|
| 6146 |
+
{
|
| 6147 |
+
"epoch": 3.59,
|
| 6148 |
+
"learning_rate": 0.0002,
|
| 6149 |
+
"loss": 0.5342,
|
| 6150 |
+
"step": 6450
|
| 6151 |
+
},
|
| 6152 |
+
{
|
| 6153 |
+
"epoch": 3.59,
|
| 6154 |
+
"learning_rate": 0.0002,
|
| 6155 |
+
"loss": 0.5037,
|
| 6156 |
+
"step": 6460
|
| 6157 |
+
},
|
| 6158 |
+
{
|
| 6159 |
+
"epoch": 3.6,
|
| 6160 |
+
"learning_rate": 0.0002,
|
| 6161 |
+
"loss": 0.5432,
|
| 6162 |
+
"step": 6470
|
| 6163 |
+
},
|
| 6164 |
+
{
|
| 6165 |
+
"epoch": 3.61,
|
| 6166 |
+
"learning_rate": 0.0002,
|
| 6167 |
+
"loss": 0.4702,
|
| 6168 |
+
"step": 6480
|
| 6169 |
+
},
|
| 6170 |
+
{
|
| 6171 |
+
"epoch": 3.61,
|
| 6172 |
+
"learning_rate": 0.0002,
|
| 6173 |
+
"loss": 0.5515,
|
| 6174 |
+
"step": 6490
|
| 6175 |
+
},
|
| 6176 |
+
{
|
| 6177 |
+
"epoch": 3.62,
|
| 6178 |
+
"learning_rate": 0.0002,
|
| 6179 |
+
"loss": 0.4315,
|
| 6180 |
+
"step": 6500
|
| 6181 |
+
},
|
| 6182 |
+
{
|
| 6183 |
+
"epoch": 3.62,
|
| 6184 |
+
"learning_rate": 0.0002,
|
| 6185 |
+
"loss": 0.4855,
|
| 6186 |
+
"step": 6510
|
| 6187 |
+
},
|
| 6188 |
+
{
|
| 6189 |
+
"epoch": 3.63,
|
| 6190 |
+
"learning_rate": 0.0002,
|
| 6191 |
+
"loss": 0.5462,
|
| 6192 |
+
"step": 6520
|
| 6193 |
+
},
|
| 6194 |
+
{
|
| 6195 |
+
"epoch": 3.63,
|
| 6196 |
+
"learning_rate": 0.0002,
|
| 6197 |
+
"loss": 0.5432,
|
| 6198 |
+
"step": 6530
|
| 6199 |
+
},
|
| 6200 |
+
{
|
| 6201 |
+
"epoch": 3.64,
|
| 6202 |
+
"learning_rate": 0.0002,
|
| 6203 |
+
"loss": 0.4896,
|
| 6204 |
+
"step": 6540
|
| 6205 |
+
},
|
| 6206 |
+
{
|
| 6207 |
+
"epoch": 3.64,
|
| 6208 |
+
"learning_rate": 0.0002,
|
| 6209 |
+
"loss": 0.5729,
|
| 6210 |
+
"step": 6550
|
| 6211 |
+
},
|
| 6212 |
+
{
|
| 6213 |
+
"epoch": 3.65,
|
| 6214 |
+
"learning_rate": 0.0002,
|
| 6215 |
+
"loss": 0.5002,
|
| 6216 |
+
"step": 6560
|
| 6217 |
+
},
|
| 6218 |
+
{
|
| 6219 |
+
"epoch": 3.66,
|
| 6220 |
+
"learning_rate": 0.0002,
|
| 6221 |
+
"loss": 0.4801,
|
| 6222 |
+
"step": 6570
|
| 6223 |
+
},
|
| 6224 |
+
{
|
| 6225 |
+
"epoch": 3.66,
|
| 6226 |
+
"learning_rate": 0.0002,
|
| 6227 |
+
"loss": 0.557,
|
| 6228 |
+
"step": 6580
|
| 6229 |
+
},
|
| 6230 |
+
{
|
| 6231 |
+
"epoch": 3.67,
|
| 6232 |
+
"learning_rate": 0.0002,
|
| 6233 |
+
"loss": 0.4869,
|
| 6234 |
+
"step": 6590
|
| 6235 |
+
},
|
| 6236 |
+
{
|
| 6237 |
+
"epoch": 3.67,
|
| 6238 |
+
"learning_rate": 0.0002,
|
| 6239 |
+
"loss": 0.5487,
|
| 6240 |
+
"step": 6600
|
| 6241 |
+
},
|
| 6242 |
+
{
|
| 6243 |
+
"epoch": 3.67,
|
| 6244 |
+
"eval_loss": 0.7634516358375549,
|
| 6245 |
+
"eval_runtime": 80.8256,
|
| 6246 |
+
"eval_samples_per_second": 12.372,
|
| 6247 |
+
"eval_steps_per_second": 6.186,
|
| 6248 |
+
"step": 6600
|
| 6249 |
+
},
|
| 6250 |
+
{
|
| 6251 |
+
"epoch": 3.67,
|
| 6252 |
+
"mmlu_eval_accuracy": 0.48499368179521124,
|
| 6253 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
|
| 6254 |
+
"mmlu_eval_accuracy_anatomy": 0.5714285714285714,
|
| 6255 |
+
"mmlu_eval_accuracy_astronomy": 0.375,
|
| 6256 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 6257 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.4482758620689655,
|
| 6258 |
+
"mmlu_eval_accuracy_college_biology": 0.375,
|
| 6259 |
+
"mmlu_eval_accuracy_college_chemistry": 0.0,
|
| 6260 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 6261 |
+
"mmlu_eval_accuracy_college_mathematics": 0.09090909090909091,
|
| 6262 |
+
"mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
|
| 6263 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 6264 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 6265 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
|
| 6266 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 6267 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.25,
|
| 6268 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
|
| 6269 |
+
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
|
| 6270 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 6271 |
+
"mmlu_eval_accuracy_high_school_biology": 0.4375,
|
| 6272 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
|
| 6273 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 6274 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 6275 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
|
| 6276 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.47619047619047616,
|
| 6277 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
|
| 6278 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
|
| 6279 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.34615384615384615,
|
| 6280 |
+
"mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
|
| 6281 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
|
| 6282 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913,
|
| 6283 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.5909090909090909,
|
| 6284 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
|
| 6285 |
+
"mmlu_eval_accuracy_human_aging": 0.6086956521739131,
|
| 6286 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 6287 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 6288 |
+
"mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
|
| 6289 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
|
| 6290 |
+
"mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
|
| 6291 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 6292 |
+
"mmlu_eval_accuracy_marketing": 0.84,
|
| 6293 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 6294 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6511627906976745,
|
| 6295 |
+
"mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
|
| 6296 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.29,
|
| 6297 |
+
"mmlu_eval_accuracy_nutrition": 0.696969696969697,
|
| 6298 |
+
"mmlu_eval_accuracy_philosophy": 0.5588235294117647,
|
| 6299 |
+
"mmlu_eval_accuracy_prehistory": 0.42857142857142855,
|
| 6300 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
|
| 6301 |
+
"mmlu_eval_accuracy_professional_law": 0.3176470588235294,
|
| 6302 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
|
| 6303 |
+
"mmlu_eval_accuracy_professional_psychology": 0.43478260869565216,
|
| 6304 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 6305 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 6306 |
+
"mmlu_eval_accuracy_sociology": 0.7272727272727273,
|
| 6307 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
|
| 6308 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 6309 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 6310 |
+
"mmlu_loss": 1.8758110864666673,
|
| 6311 |
+
"step": 6600
|
| 6312 |
+
},
|
| 6313 |
+
{
|
| 6314 |
+
"epoch": 3.68,
|
| 6315 |
+
"learning_rate": 0.0002,
|
| 6316 |
+
"loss": 0.5892,
|
| 6317 |
+
"step": 6610
|
| 6318 |
+
},
|
| 6319 |
+
{
|
| 6320 |
+
"epoch": 3.68,
|
| 6321 |
+
"learning_rate": 0.0002,
|
| 6322 |
+
"loss": 0.4989,
|
| 6323 |
+
"step": 6620
|
| 6324 |
+
},
|
| 6325 |
+
{
|
| 6326 |
+
"epoch": 3.69,
|
| 6327 |
+
"learning_rate": 0.0002,
|
| 6328 |
+
"loss": 0.609,
|
| 6329 |
+
"step": 6630
|
| 6330 |
+
},
|
| 6331 |
+
{
|
| 6332 |
+
"epoch": 3.7,
|
| 6333 |
+
"learning_rate": 0.0002,
|
| 6334 |
+
"loss": 0.4694,
|
| 6335 |
+
"step": 6640
|
| 6336 |
+
},
|
| 6337 |
+
{
|
| 6338 |
+
"epoch": 3.7,
|
| 6339 |
+
"learning_rate": 0.0002,
|
| 6340 |
+
"loss": 0.5057,
|
| 6341 |
+
"step": 6650
|
| 6342 |
+
},
|
| 6343 |
+
{
|
| 6344 |
+
"epoch": 3.71,
|
| 6345 |
+
"learning_rate": 0.0002,
|
| 6346 |
+
"loss": 0.4766,
|
| 6347 |
+
"step": 6660
|
| 6348 |
+
},
|
| 6349 |
+
{
|
| 6350 |
+
"epoch": 3.71,
|
| 6351 |
+
"learning_rate": 0.0002,
|
| 6352 |
+
"loss": 0.5479,
|
| 6353 |
+
"step": 6670
|
| 6354 |
+
},
|
| 6355 |
+
{
|
| 6356 |
+
"epoch": 3.72,
|
| 6357 |
+
"learning_rate": 0.0002,
|
| 6358 |
+
"loss": 0.5301,
|
| 6359 |
+
"step": 6680
|
| 6360 |
+
},
|
| 6361 |
+
{
|
| 6362 |
+
"epoch": 3.72,
|
| 6363 |
+
"learning_rate": 0.0002,
|
| 6364 |
+
"loss": 0.4815,
|
| 6365 |
+
"step": 6690
|
| 6366 |
+
},
|
| 6367 |
+
{
|
| 6368 |
+
"epoch": 3.73,
|
| 6369 |
+
"learning_rate": 0.0002,
|
| 6370 |
+
"loss": 0.5403,
|
| 6371 |
+
"step": 6700
|
| 6372 |
+
},
|
| 6373 |
+
{
|
| 6374 |
+
"epoch": 3.73,
|
| 6375 |
+
"learning_rate": 0.0002,
|
| 6376 |
+
"loss": 0.479,
|
| 6377 |
+
"step": 6710
|
| 6378 |
+
},
|
| 6379 |
+
{
|
| 6380 |
+
"epoch": 3.74,
|
| 6381 |
+
"learning_rate": 0.0002,
|
| 6382 |
+
"loss": 0.527,
|
| 6383 |
+
"step": 6720
|
| 6384 |
+
},
|
| 6385 |
+
{
|
| 6386 |
+
"epoch": 3.75,
|
| 6387 |
+
"learning_rate": 0.0002,
|
| 6388 |
+
"loss": 0.5906,
|
| 6389 |
+
"step": 6730
|
| 6390 |
+
},
|
| 6391 |
+
{
|
| 6392 |
+
"epoch": 3.75,
|
| 6393 |
+
"learning_rate": 0.0002,
|
| 6394 |
+
"loss": 0.4823,
|
| 6395 |
+
"step": 6740
|
| 6396 |
+
},
|
| 6397 |
+
{
|
| 6398 |
+
"epoch": 3.76,
|
| 6399 |
+
"learning_rate": 0.0002,
|
| 6400 |
+
"loss": 0.5188,
|
| 6401 |
+
"step": 6750
|
| 6402 |
+
},
|
| 6403 |
+
{
|
| 6404 |
+
"epoch": 3.76,
|
| 6405 |
+
"learning_rate": 0.0002,
|
| 6406 |
+
"loss": 0.498,
|
| 6407 |
+
"step": 6760
|
| 6408 |
+
},
|
| 6409 |
+
{
|
| 6410 |
+
"epoch": 3.77,
|
| 6411 |
+
"learning_rate": 0.0002,
|
| 6412 |
+
"loss": 0.4929,
|
| 6413 |
+
"step": 6770
|
| 6414 |
+
},
|
| 6415 |
+
{
|
| 6416 |
+
"epoch": 3.77,
|
| 6417 |
+
"learning_rate": 0.0002,
|
| 6418 |
+
"loss": 0.4363,
|
| 6419 |
+
"step": 6780
|
| 6420 |
+
},
|
| 6421 |
+
{
|
| 6422 |
+
"epoch": 3.78,
|
| 6423 |
+
"learning_rate": 0.0002,
|
| 6424 |
+
"loss": 0.4736,
|
| 6425 |
+
"step": 6790
|
| 6426 |
+
},
|
| 6427 |
+
{
|
| 6428 |
+
"epoch": 3.78,
|
| 6429 |
+
"learning_rate": 0.0002,
|
| 6430 |
+
"loss": 0.5008,
|
| 6431 |
+
"step": 6800
|
| 6432 |
+
},
|
| 6433 |
+
{
|
| 6434 |
+
"epoch": 3.78,
|
| 6435 |
+
"eval_loss": 0.7794287204742432,
|
| 6436 |
+
"eval_runtime": 80.8425,
|
| 6437 |
+
"eval_samples_per_second": 12.37,
|
| 6438 |
+
"eval_steps_per_second": 6.185,
|
| 6439 |
+
"step": 6800
|
| 6440 |
+
},
|
| 6441 |
+
{
|
| 6442 |
+
"epoch": 3.78,
|
| 6443 |
+
"mmlu_eval_accuracy": 0.47885377305881066,
|
| 6444 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
|
| 6445 |
+
"mmlu_eval_accuracy_anatomy": 0.5714285714285714,
|
| 6446 |
+
"mmlu_eval_accuracy_astronomy": 0.4375,
|
| 6447 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 6448 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.41379310344827586,
|
| 6449 |
+
"mmlu_eval_accuracy_college_biology": 0.5,
|
| 6450 |
+
"mmlu_eval_accuracy_college_chemistry": 0.0,
|
| 6451 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 6452 |
+
"mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
|
| 6453 |
+
"mmlu_eval_accuracy_college_medicine": 0.36363636363636365,
|
| 6454 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 6455 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 6456 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 6457 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 6458 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.25,
|
| 6459 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.2926829268292683,
|
| 6460 |
+
"mmlu_eval_accuracy_formal_logic": 0.07142857142857142,
|
| 6461 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 6462 |
+
"mmlu_eval_accuracy_high_school_biology": 0.4375,
|
| 6463 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
|
| 6464 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 6465 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 6466 |
+
"mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
|
| 6467 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5238095238095238,
|
| 6468 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
|
| 6469 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
|
| 6470 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.3076923076923077,
|
| 6471 |
+
"mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
|
| 6472 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8,
|
| 6473 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.21739130434782608,
|
| 6474 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.5909090909090909,
|
| 6475 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
|
| 6476 |
+
"mmlu_eval_accuracy_human_aging": 0.6086956521739131,
|
| 6477 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 6478 |
+
"mmlu_eval_accuracy_international_law": 0.7692307692307693,
|
| 6479 |
+
"mmlu_eval_accuracy_jurisprudence": 0.2727272727272727,
|
| 6480 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
|
| 6481 |
+
"mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
|
| 6482 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 6483 |
+
"mmlu_eval_accuracy_marketing": 0.84,
|
| 6484 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 6485 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
|
| 6486 |
+
"mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
|
| 6487 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.24,
|
| 6488 |
+
"mmlu_eval_accuracy_nutrition": 0.696969696969697,
|
| 6489 |
+
"mmlu_eval_accuracy_philosophy": 0.5294117647058824,
|
| 6490 |
+
"mmlu_eval_accuracy_prehistory": 0.4857142857142857,
|
| 6491 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
|
| 6492 |
+
"mmlu_eval_accuracy_professional_law": 0.35294117647058826,
|
| 6493 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
|
| 6494 |
+
"mmlu_eval_accuracy_professional_psychology": 0.43478260869565216,
|
| 6495 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 6496 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 6497 |
+
"mmlu_eval_accuracy_sociology": 0.7272727272727273,
|
| 6498 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
|
| 6499 |
+
"mmlu_eval_accuracy_virology": 0.5,
|
| 6500 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 6501 |
+
"mmlu_loss": 1.9438237615182568,
|
| 6502 |
+
"step": 6800
|
| 6503 |
}
|
| 6504 |
],
|
| 6505 |
"max_steps": 10000,
|
| 6506 |
"num_train_epochs": 6,
|
| 6507 |
+
"total_flos": 3.306490728912814e+17,
|
| 6508 |
"trial_name": null,
|
| 6509 |
"trial_params": null
|
| 6510 |
}
|
{checkpoint-4800 → checkpoint-6800}/training_args.bin
RENAMED
|
File without changes
|