Farouk commited on
Commit ·
9cdf46f
1
Parent(s): 8e435ca
Training in progress, step 6200
Browse files- adapter_model.bin +1 -1
- checkpoint-4000/adapter_model/adapter_model/README.md +12 -0
- checkpoint-4000/adapter_model/adapter_model/adapter_model.bin +1 -1
- {checkpoint-4200 → checkpoint-6200}/README.md +0 -0
- {checkpoint-4200 → checkpoint-6200}/adapter_config.json +0 -0
- {checkpoint-4200 → checkpoint-6200}/adapter_model.bin +1 -1
- {checkpoint-4200 → checkpoint-6200}/added_tokens.json +0 -0
- {checkpoint-4200 → checkpoint-6200}/optimizer.pt +1 -1
- {checkpoint-4200 → checkpoint-6200}/rng_state.pth +1 -1
- {checkpoint-4200 → checkpoint-6200}/scheduler.pt +1 -1
- {checkpoint-4200 → checkpoint-6200}/special_tokens_map.json +0 -0
- {checkpoint-4200 → checkpoint-6200}/tokenizer.model +0 -0
- {checkpoint-4200 → checkpoint-6200}/tokenizer_config.json +0 -0
- {checkpoint-4200 → checkpoint-6200}/trainer_state.json +1913 -3
- {checkpoint-4200 → checkpoint-6200}/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:b46dfbc97b26476d5cc53984e35d2e29d4f127428f50303827f2d18321bd98c7
|
| 3 |
size 319977229
|
checkpoint-4000/adapter_model/adapter_model/README.md
CHANGED
|
@@ -103,6 +103,17 @@ The following `bitsandbytes` quantization config was used during training:
|
|
| 103 |
- bnb_4bit_use_double_quant: True
|
| 104 |
- bnb_4bit_compute_dtype: bfloat16
|
| 105 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 106 |
The following `bitsandbytes` quantization config was used during training:
|
| 107 |
- load_in_8bit: False
|
| 108 |
- load_in_4bit: True
|
|
@@ -124,5 +135,6 @@ The following `bitsandbytes` quantization config was used during training:
|
|
| 124 |
- PEFT 0.4.0
|
| 125 |
- PEFT 0.4.0
|
| 126 |
- PEFT 0.4.0
|
|
|
|
| 127 |
|
| 128 |
- PEFT 0.4.0
|
|
|
|
| 103 |
- bnb_4bit_use_double_quant: True
|
| 104 |
- bnb_4bit_compute_dtype: bfloat16
|
| 105 |
|
| 106 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 107 |
+
- load_in_8bit: False
|
| 108 |
+
- load_in_4bit: True
|
| 109 |
+
- llm_int8_threshold: 6.0
|
| 110 |
+
- llm_int8_skip_modules: None
|
| 111 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 112 |
+
- llm_int8_has_fp16_weight: False
|
| 113 |
+
- bnb_4bit_quant_type: nf4
|
| 114 |
+
- bnb_4bit_use_double_quant: True
|
| 115 |
+
- bnb_4bit_compute_dtype: bfloat16
|
| 116 |
+
|
| 117 |
The following `bitsandbytes` quantization config was used during training:
|
| 118 |
- load_in_8bit: False
|
| 119 |
- load_in_4bit: True
|
|
|
|
| 135 |
- PEFT 0.4.0
|
| 136 |
- PEFT 0.4.0
|
| 137 |
- PEFT 0.4.0
|
| 138 |
+
- PEFT 0.4.0
|
| 139 |
|
| 140 |
- PEFT 0.4.0
|
checkpoint-4000/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 sha256:9ea76527e5189725064e55dd22c30be194cb090546d57d65d87086d49947036a
|
| 3 |
size 319977229
|
{checkpoint-4200 → checkpoint-6200}/README.md
RENAMED
|
File without changes
|
{checkpoint-4200 → checkpoint-6200}/adapter_config.json
RENAMED
|
File without changes
|
{checkpoint-4200 → checkpoint-6200}/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:b46dfbc97b26476d5cc53984e35d2e29d4f127428f50303827f2d18321bd98c7
|
| 3 |
size 319977229
|
{checkpoint-4200 → checkpoint-6200}/added_tokens.json
RENAMED
|
File without changes
|
{checkpoint-4200 → checkpoint-6200}/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:a0773fe4117707e0ae2c5b23415323a7f2b7162119103a5e880819e8e7ea8f77
|
| 3 |
size 1279539973
|
{checkpoint-4200 → checkpoint-6200}/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:ad84a667992a8ba1d3d2486c993cc7084110e1996db092f8e4674053b0643fa9
|
| 3 |
size 14511
|
{checkpoint-4200 → checkpoint-6200}/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 sha256:00a2cdb3e0cba9c4eb9fe8ce4b971cf52e18a108de1c45e1a56c83cbbb819ccc
|
| 3 |
size 627
|
{checkpoint-4200 → checkpoint-6200}/special_tokens_map.json
RENAMED
|
File without changes
|
{checkpoint-4200 → checkpoint-6200}/tokenizer.model
RENAMED
|
File without changes
|
{checkpoint-4200 → checkpoint-6200}/tokenizer_config.json
RENAMED
|
File without changes
|
{checkpoint-4200 → checkpoint-6200}/trainer_state.json
RENAMED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": 0.758108913898468,
|
| 3 |
"best_model_checkpoint": "experts/expert-5/checkpoint-4000",
|
| 4 |
-
"epoch": 1.
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -4017,11 +4017,1921 @@
|
|
| 4017 |
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 4018 |
"mmlu_loss": 1.3325689750536303,
|
| 4019 |
"step": 4200
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4020 |
}
|
| 4021 |
],
|
| 4022 |
"max_steps": 10000,
|
| 4023 |
"num_train_epochs": 3,
|
| 4024 |
-
"total_flos":
|
| 4025 |
"trial_name": null,
|
| 4026 |
"trial_params": null
|
| 4027 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": 0.758108913898468,
|
| 3 |
"best_model_checkpoint": "experts/expert-5/checkpoint-4000",
|
| 4 |
+
"epoch": 1.500484027105518,
|
| 5 |
+
"global_step": 6200,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 4017 |
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 4018 |
"mmlu_loss": 1.3325689750536303,
|
| 4019 |
"step": 4200
|
| 4020 |
+
},
|
| 4021 |
+
{
|
| 4022 |
+
"epoch": 1.02,
|
| 4023 |
+
"learning_rate": 0.0002,
|
| 4024 |
+
"loss": 0.6978,
|
| 4025 |
+
"step": 4210
|
| 4026 |
+
},
|
| 4027 |
+
{
|
| 4028 |
+
"epoch": 1.02,
|
| 4029 |
+
"learning_rate": 0.0002,
|
| 4030 |
+
"loss": 0.6938,
|
| 4031 |
+
"step": 4220
|
| 4032 |
+
},
|
| 4033 |
+
{
|
| 4034 |
+
"epoch": 1.02,
|
| 4035 |
+
"learning_rate": 0.0002,
|
| 4036 |
+
"loss": 0.7215,
|
| 4037 |
+
"step": 4230
|
| 4038 |
+
},
|
| 4039 |
+
{
|
| 4040 |
+
"epoch": 1.03,
|
| 4041 |
+
"learning_rate": 0.0002,
|
| 4042 |
+
"loss": 0.6615,
|
| 4043 |
+
"step": 4240
|
| 4044 |
+
},
|
| 4045 |
+
{
|
| 4046 |
+
"epoch": 1.03,
|
| 4047 |
+
"learning_rate": 0.0002,
|
| 4048 |
+
"loss": 0.6689,
|
| 4049 |
+
"step": 4250
|
| 4050 |
+
},
|
| 4051 |
+
{
|
| 4052 |
+
"epoch": 1.03,
|
| 4053 |
+
"learning_rate": 0.0002,
|
| 4054 |
+
"loss": 0.6978,
|
| 4055 |
+
"step": 4260
|
| 4056 |
+
},
|
| 4057 |
+
{
|
| 4058 |
+
"epoch": 1.03,
|
| 4059 |
+
"learning_rate": 0.0002,
|
| 4060 |
+
"loss": 0.7192,
|
| 4061 |
+
"step": 4270
|
| 4062 |
+
},
|
| 4063 |
+
{
|
| 4064 |
+
"epoch": 1.04,
|
| 4065 |
+
"learning_rate": 0.0002,
|
| 4066 |
+
"loss": 0.6788,
|
| 4067 |
+
"step": 4280
|
| 4068 |
+
},
|
| 4069 |
+
{
|
| 4070 |
+
"epoch": 1.04,
|
| 4071 |
+
"learning_rate": 0.0002,
|
| 4072 |
+
"loss": 0.7332,
|
| 4073 |
+
"step": 4290
|
| 4074 |
+
},
|
| 4075 |
+
{
|
| 4076 |
+
"epoch": 1.04,
|
| 4077 |
+
"learning_rate": 0.0002,
|
| 4078 |
+
"loss": 0.6555,
|
| 4079 |
+
"step": 4300
|
| 4080 |
+
},
|
| 4081 |
+
{
|
| 4082 |
+
"epoch": 1.04,
|
| 4083 |
+
"learning_rate": 0.0002,
|
| 4084 |
+
"loss": 0.6682,
|
| 4085 |
+
"step": 4310
|
| 4086 |
+
},
|
| 4087 |
+
{
|
| 4088 |
+
"epoch": 1.05,
|
| 4089 |
+
"learning_rate": 0.0002,
|
| 4090 |
+
"loss": 0.6731,
|
| 4091 |
+
"step": 4320
|
| 4092 |
+
},
|
| 4093 |
+
{
|
| 4094 |
+
"epoch": 1.05,
|
| 4095 |
+
"learning_rate": 0.0002,
|
| 4096 |
+
"loss": 0.6504,
|
| 4097 |
+
"step": 4330
|
| 4098 |
+
},
|
| 4099 |
+
{
|
| 4100 |
+
"epoch": 1.05,
|
| 4101 |
+
"learning_rate": 0.0002,
|
| 4102 |
+
"loss": 0.6379,
|
| 4103 |
+
"step": 4340
|
| 4104 |
+
},
|
| 4105 |
+
{
|
| 4106 |
+
"epoch": 1.05,
|
| 4107 |
+
"learning_rate": 0.0002,
|
| 4108 |
+
"loss": 0.6272,
|
| 4109 |
+
"step": 4350
|
| 4110 |
+
},
|
| 4111 |
+
{
|
| 4112 |
+
"epoch": 1.06,
|
| 4113 |
+
"learning_rate": 0.0002,
|
| 4114 |
+
"loss": 0.599,
|
| 4115 |
+
"step": 4360
|
| 4116 |
+
},
|
| 4117 |
+
{
|
| 4118 |
+
"epoch": 1.06,
|
| 4119 |
+
"learning_rate": 0.0002,
|
| 4120 |
+
"loss": 0.6497,
|
| 4121 |
+
"step": 4370
|
| 4122 |
+
},
|
| 4123 |
+
{
|
| 4124 |
+
"epoch": 1.06,
|
| 4125 |
+
"learning_rate": 0.0002,
|
| 4126 |
+
"loss": 0.6926,
|
| 4127 |
+
"step": 4380
|
| 4128 |
+
},
|
| 4129 |
+
{
|
| 4130 |
+
"epoch": 1.06,
|
| 4131 |
+
"learning_rate": 0.0002,
|
| 4132 |
+
"loss": 0.6493,
|
| 4133 |
+
"step": 4390
|
| 4134 |
+
},
|
| 4135 |
+
{
|
| 4136 |
+
"epoch": 1.06,
|
| 4137 |
+
"learning_rate": 0.0002,
|
| 4138 |
+
"loss": 0.6835,
|
| 4139 |
+
"step": 4400
|
| 4140 |
+
},
|
| 4141 |
+
{
|
| 4142 |
+
"epoch": 1.06,
|
| 4143 |
+
"eval_loss": 0.7639644742012024,
|
| 4144 |
+
"eval_runtime": 146.9269,
|
| 4145 |
+
"eval_samples_per_second": 6.806,
|
| 4146 |
+
"eval_steps_per_second": 3.403,
|
| 4147 |
+
"step": 4400
|
| 4148 |
+
},
|
| 4149 |
+
{
|
| 4150 |
+
"epoch": 1.06,
|
| 4151 |
+
"mmlu_eval_accuracy": 0.4935246375389956,
|
| 4152 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
|
| 4153 |
+
"mmlu_eval_accuracy_anatomy": 0.7857142857142857,
|
| 4154 |
+
"mmlu_eval_accuracy_astronomy": 0.4375,
|
| 4155 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 4156 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
|
| 4157 |
+
"mmlu_eval_accuracy_college_biology": 0.375,
|
| 4158 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 4159 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 4160 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 4161 |
+
"mmlu_eval_accuracy_college_medicine": 0.36363636363636365,
|
| 4162 |
+
"mmlu_eval_accuracy_college_physics": 0.2727272727272727,
|
| 4163 |
+
"mmlu_eval_accuracy_computer_security": 0.5454545454545454,
|
| 4164 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 4165 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 4166 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.25,
|
| 4167 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 4168 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 4169 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 4170 |
+
"mmlu_eval_accuracy_high_school_biology": 0.5,
|
| 4171 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.22727272727272727,
|
| 4172 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 4173 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 4174 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
|
| 4175 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5238095238095238,
|
| 4176 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4883720930232558,
|
| 4177 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
|
| 4178 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.5,
|
| 4179 |
+
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
|
| 4180 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
|
| 4181 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
|
| 4182 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
|
| 4183 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
|
| 4184 |
+
"mmlu_eval_accuracy_human_aging": 0.7391304347826086,
|
| 4185 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 4186 |
+
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
|
| 4187 |
+
"mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
|
| 4188 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
|
| 4189 |
+
"mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
|
| 4190 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 4191 |
+
"mmlu_eval_accuracy_marketing": 0.76,
|
| 4192 |
+
"mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
|
| 4193 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
|
| 4194 |
+
"mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
|
| 4195 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.23,
|
| 4196 |
+
"mmlu_eval_accuracy_nutrition": 0.6666666666666666,
|
| 4197 |
+
"mmlu_eval_accuracy_philosophy": 0.5,
|
| 4198 |
+
"mmlu_eval_accuracy_prehistory": 0.6,
|
| 4199 |
+
"mmlu_eval_accuracy_professional_accounting": 0.2903225806451613,
|
| 4200 |
+
"mmlu_eval_accuracy_professional_law": 0.34705882352941175,
|
| 4201 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
|
| 4202 |
+
"mmlu_eval_accuracy_professional_psychology": 0.5072463768115942,
|
| 4203 |
+
"mmlu_eval_accuracy_public_relations": 0.5833333333333334,
|
| 4204 |
+
"mmlu_eval_accuracy_security_studies": 0.5185185185185185,
|
| 4205 |
+
"mmlu_eval_accuracy_sociology": 0.6818181818181818,
|
| 4206 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
|
| 4207 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 4208 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 4209 |
+
"mmlu_loss": 1.2368104296393556,
|
| 4210 |
+
"step": 4400
|
| 4211 |
+
},
|
| 4212 |
+
{
|
| 4213 |
+
"epoch": 1.07,
|
| 4214 |
+
"learning_rate": 0.0002,
|
| 4215 |
+
"loss": 0.6258,
|
| 4216 |
+
"step": 4410
|
| 4217 |
+
},
|
| 4218 |
+
{
|
| 4219 |
+
"epoch": 1.07,
|
| 4220 |
+
"learning_rate": 0.0002,
|
| 4221 |
+
"loss": 0.6949,
|
| 4222 |
+
"step": 4420
|
| 4223 |
+
},
|
| 4224 |
+
{
|
| 4225 |
+
"epoch": 1.07,
|
| 4226 |
+
"learning_rate": 0.0002,
|
| 4227 |
+
"loss": 0.7505,
|
| 4228 |
+
"step": 4430
|
| 4229 |
+
},
|
| 4230 |
+
{
|
| 4231 |
+
"epoch": 1.07,
|
| 4232 |
+
"learning_rate": 0.0002,
|
| 4233 |
+
"loss": 0.6799,
|
| 4234 |
+
"step": 4440
|
| 4235 |
+
},
|
| 4236 |
+
{
|
| 4237 |
+
"epoch": 1.08,
|
| 4238 |
+
"learning_rate": 0.0002,
|
| 4239 |
+
"loss": 0.6257,
|
| 4240 |
+
"step": 4450
|
| 4241 |
+
},
|
| 4242 |
+
{
|
| 4243 |
+
"epoch": 1.08,
|
| 4244 |
+
"learning_rate": 0.0002,
|
| 4245 |
+
"loss": 0.6309,
|
| 4246 |
+
"step": 4460
|
| 4247 |
+
},
|
| 4248 |
+
{
|
| 4249 |
+
"epoch": 1.08,
|
| 4250 |
+
"learning_rate": 0.0002,
|
| 4251 |
+
"loss": 0.7105,
|
| 4252 |
+
"step": 4470
|
| 4253 |
+
},
|
| 4254 |
+
{
|
| 4255 |
+
"epoch": 1.08,
|
| 4256 |
+
"learning_rate": 0.0002,
|
| 4257 |
+
"loss": 0.6771,
|
| 4258 |
+
"step": 4480
|
| 4259 |
+
},
|
| 4260 |
+
{
|
| 4261 |
+
"epoch": 1.09,
|
| 4262 |
+
"learning_rate": 0.0002,
|
| 4263 |
+
"loss": 0.663,
|
| 4264 |
+
"step": 4490
|
| 4265 |
+
},
|
| 4266 |
+
{
|
| 4267 |
+
"epoch": 1.09,
|
| 4268 |
+
"learning_rate": 0.0002,
|
| 4269 |
+
"loss": 0.6604,
|
| 4270 |
+
"step": 4500
|
| 4271 |
+
},
|
| 4272 |
+
{
|
| 4273 |
+
"epoch": 1.09,
|
| 4274 |
+
"learning_rate": 0.0002,
|
| 4275 |
+
"loss": 0.6141,
|
| 4276 |
+
"step": 4510
|
| 4277 |
+
},
|
| 4278 |
+
{
|
| 4279 |
+
"epoch": 1.09,
|
| 4280 |
+
"learning_rate": 0.0002,
|
| 4281 |
+
"loss": 0.6747,
|
| 4282 |
+
"step": 4520
|
| 4283 |
+
},
|
| 4284 |
+
{
|
| 4285 |
+
"epoch": 1.1,
|
| 4286 |
+
"learning_rate": 0.0002,
|
| 4287 |
+
"loss": 0.6782,
|
| 4288 |
+
"step": 4530
|
| 4289 |
+
},
|
| 4290 |
+
{
|
| 4291 |
+
"epoch": 1.1,
|
| 4292 |
+
"learning_rate": 0.0002,
|
| 4293 |
+
"loss": 0.7018,
|
| 4294 |
+
"step": 4540
|
| 4295 |
+
},
|
| 4296 |
+
{
|
| 4297 |
+
"epoch": 1.1,
|
| 4298 |
+
"learning_rate": 0.0002,
|
| 4299 |
+
"loss": 0.6919,
|
| 4300 |
+
"step": 4550
|
| 4301 |
+
},
|
| 4302 |
+
{
|
| 4303 |
+
"epoch": 1.1,
|
| 4304 |
+
"learning_rate": 0.0002,
|
| 4305 |
+
"loss": 0.6913,
|
| 4306 |
+
"step": 4560
|
| 4307 |
+
},
|
| 4308 |
+
{
|
| 4309 |
+
"epoch": 1.11,
|
| 4310 |
+
"learning_rate": 0.0002,
|
| 4311 |
+
"loss": 0.7025,
|
| 4312 |
+
"step": 4570
|
| 4313 |
+
},
|
| 4314 |
+
{
|
| 4315 |
+
"epoch": 1.11,
|
| 4316 |
+
"learning_rate": 0.0002,
|
| 4317 |
+
"loss": 0.688,
|
| 4318 |
+
"step": 4580
|
| 4319 |
+
},
|
| 4320 |
+
{
|
| 4321 |
+
"epoch": 1.11,
|
| 4322 |
+
"learning_rate": 0.0002,
|
| 4323 |
+
"loss": 0.6868,
|
| 4324 |
+
"step": 4590
|
| 4325 |
+
},
|
| 4326 |
+
{
|
| 4327 |
+
"epoch": 1.11,
|
| 4328 |
+
"learning_rate": 0.0002,
|
| 4329 |
+
"loss": 0.6542,
|
| 4330 |
+
"step": 4600
|
| 4331 |
+
},
|
| 4332 |
+
{
|
| 4333 |
+
"epoch": 1.11,
|
| 4334 |
+
"eval_loss": 0.767784059047699,
|
| 4335 |
+
"eval_runtime": 146.9032,
|
| 4336 |
+
"eval_samples_per_second": 6.807,
|
| 4337 |
+
"eval_steps_per_second": 3.404,
|
| 4338 |
+
"step": 4600
|
| 4339 |
+
},
|
| 4340 |
+
{
|
| 4341 |
+
"epoch": 1.11,
|
| 4342 |
+
"mmlu_eval_accuracy": 0.4969588552133961,
|
| 4343 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
|
| 4344 |
+
"mmlu_eval_accuracy_anatomy": 0.7142857142857143,
|
| 4345 |
+
"mmlu_eval_accuracy_astronomy": 0.4375,
|
| 4346 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 4347 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
|
| 4348 |
+
"mmlu_eval_accuracy_college_biology": 0.375,
|
| 4349 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 4350 |
+
"mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
|
| 4351 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 4352 |
+
"mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
|
| 4353 |
+
"mmlu_eval_accuracy_college_physics": 0.36363636363636365,
|
| 4354 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 4355 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 4356 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 4357 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.25,
|
| 4358 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 4359 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 4360 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 4361 |
+
"mmlu_eval_accuracy_high_school_biology": 0.5,
|
| 4362 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
|
| 4363 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 4364 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 4365 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
|
| 4366 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5238095238095238,
|
| 4367 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.46511627906976744,
|
| 4368 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
|
| 4369 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
|
| 4370 |
+
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
|
| 4371 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333,
|
| 4372 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
|
| 4373 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
|
| 4374 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
|
| 4375 |
+
"mmlu_eval_accuracy_human_aging": 0.6521739130434783,
|
| 4376 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 4377 |
+
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
|
| 4378 |
+
"mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
|
| 4379 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
|
| 4380 |
+
"mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
|
| 4381 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 4382 |
+
"mmlu_eval_accuracy_marketing": 0.76,
|
| 4383 |
+
"mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
|
| 4384 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
|
| 4385 |
+
"mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
|
| 4386 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.24,
|
| 4387 |
+
"mmlu_eval_accuracy_nutrition": 0.696969696969697,
|
| 4388 |
+
"mmlu_eval_accuracy_philosophy": 0.5294117647058824,
|
| 4389 |
+
"mmlu_eval_accuracy_prehistory": 0.5428571428571428,
|
| 4390 |
+
"mmlu_eval_accuracy_professional_accounting": 0.2903225806451613,
|
| 4391 |
+
"mmlu_eval_accuracy_professional_law": 0.36470588235294116,
|
| 4392 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
|
| 4393 |
+
"mmlu_eval_accuracy_professional_psychology": 0.5072463768115942,
|
| 4394 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 4395 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 4396 |
+
"mmlu_eval_accuracy_sociology": 0.7727272727272727,
|
| 4397 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 4398 |
+
"mmlu_eval_accuracy_virology": 0.5,
|
| 4399 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 4400 |
+
"mmlu_loss": 1.1984694695036966,
|
| 4401 |
+
"step": 4600
|
| 4402 |
+
},
|
| 4403 |
+
{
|
| 4404 |
+
"epoch": 1.12,
|
| 4405 |
+
"learning_rate": 0.0002,
|
| 4406 |
+
"loss": 0.7047,
|
| 4407 |
+
"step": 4610
|
| 4408 |
+
},
|
| 4409 |
+
{
|
| 4410 |
+
"epoch": 1.12,
|
| 4411 |
+
"learning_rate": 0.0002,
|
| 4412 |
+
"loss": 0.691,
|
| 4413 |
+
"step": 4620
|
| 4414 |
+
},
|
| 4415 |
+
{
|
| 4416 |
+
"epoch": 1.12,
|
| 4417 |
+
"learning_rate": 0.0002,
|
| 4418 |
+
"loss": 0.8366,
|
| 4419 |
+
"step": 4630
|
| 4420 |
+
},
|
| 4421 |
+
{
|
| 4422 |
+
"epoch": 1.12,
|
| 4423 |
+
"learning_rate": 0.0002,
|
| 4424 |
+
"loss": 0.7076,
|
| 4425 |
+
"step": 4640
|
| 4426 |
+
},
|
| 4427 |
+
{
|
| 4428 |
+
"epoch": 1.13,
|
| 4429 |
+
"learning_rate": 0.0002,
|
| 4430 |
+
"loss": 0.6524,
|
| 4431 |
+
"step": 4650
|
| 4432 |
+
},
|
| 4433 |
+
{
|
| 4434 |
+
"epoch": 1.13,
|
| 4435 |
+
"learning_rate": 0.0002,
|
| 4436 |
+
"loss": 0.758,
|
| 4437 |
+
"step": 4660
|
| 4438 |
+
},
|
| 4439 |
+
{
|
| 4440 |
+
"epoch": 1.13,
|
| 4441 |
+
"learning_rate": 0.0002,
|
| 4442 |
+
"loss": 0.7079,
|
| 4443 |
+
"step": 4670
|
| 4444 |
+
},
|
| 4445 |
+
{
|
| 4446 |
+
"epoch": 1.13,
|
| 4447 |
+
"learning_rate": 0.0002,
|
| 4448 |
+
"loss": 0.7079,
|
| 4449 |
+
"step": 4680
|
| 4450 |
+
},
|
| 4451 |
+
{
|
| 4452 |
+
"epoch": 1.14,
|
| 4453 |
+
"learning_rate": 0.0002,
|
| 4454 |
+
"loss": 0.6862,
|
| 4455 |
+
"step": 4690
|
| 4456 |
+
},
|
| 4457 |
+
{
|
| 4458 |
+
"epoch": 1.14,
|
| 4459 |
+
"learning_rate": 0.0002,
|
| 4460 |
+
"loss": 0.5953,
|
| 4461 |
+
"step": 4700
|
| 4462 |
+
},
|
| 4463 |
+
{
|
| 4464 |
+
"epoch": 1.14,
|
| 4465 |
+
"learning_rate": 0.0002,
|
| 4466 |
+
"loss": 0.707,
|
| 4467 |
+
"step": 4710
|
| 4468 |
+
},
|
| 4469 |
+
{
|
| 4470 |
+
"epoch": 1.14,
|
| 4471 |
+
"learning_rate": 0.0002,
|
| 4472 |
+
"loss": 0.6508,
|
| 4473 |
+
"step": 4720
|
| 4474 |
+
},
|
| 4475 |
+
{
|
| 4476 |
+
"epoch": 1.14,
|
| 4477 |
+
"learning_rate": 0.0002,
|
| 4478 |
+
"loss": 0.6764,
|
| 4479 |
+
"step": 4730
|
| 4480 |
+
},
|
| 4481 |
+
{
|
| 4482 |
+
"epoch": 1.15,
|
| 4483 |
+
"learning_rate": 0.0002,
|
| 4484 |
+
"loss": 0.6385,
|
| 4485 |
+
"step": 4740
|
| 4486 |
+
},
|
| 4487 |
+
{
|
| 4488 |
+
"epoch": 1.15,
|
| 4489 |
+
"learning_rate": 0.0002,
|
| 4490 |
+
"loss": 0.7513,
|
| 4491 |
+
"step": 4750
|
| 4492 |
+
},
|
| 4493 |
+
{
|
| 4494 |
+
"epoch": 1.15,
|
| 4495 |
+
"learning_rate": 0.0002,
|
| 4496 |
+
"loss": 0.7166,
|
| 4497 |
+
"step": 4760
|
| 4498 |
+
},
|
| 4499 |
+
{
|
| 4500 |
+
"epoch": 1.15,
|
| 4501 |
+
"learning_rate": 0.0002,
|
| 4502 |
+
"loss": 0.6637,
|
| 4503 |
+
"step": 4770
|
| 4504 |
+
},
|
| 4505 |
+
{
|
| 4506 |
+
"epoch": 1.16,
|
| 4507 |
+
"learning_rate": 0.0002,
|
| 4508 |
+
"loss": 0.67,
|
| 4509 |
+
"step": 4780
|
| 4510 |
+
},
|
| 4511 |
+
{
|
| 4512 |
+
"epoch": 1.16,
|
| 4513 |
+
"learning_rate": 0.0002,
|
| 4514 |
+
"loss": 0.6138,
|
| 4515 |
+
"step": 4790
|
| 4516 |
+
},
|
| 4517 |
+
{
|
| 4518 |
+
"epoch": 1.16,
|
| 4519 |
+
"learning_rate": 0.0002,
|
| 4520 |
+
"loss": 0.676,
|
| 4521 |
+
"step": 4800
|
| 4522 |
+
},
|
| 4523 |
+
{
|
| 4524 |
+
"epoch": 1.16,
|
| 4525 |
+
"eval_loss": 0.7651529312133789,
|
| 4526 |
+
"eval_runtime": 147.0325,
|
| 4527 |
+
"eval_samples_per_second": 6.801,
|
| 4528 |
+
"eval_steps_per_second": 3.401,
|
| 4529 |
+
"step": 4800
|
| 4530 |
+
},
|
| 4531 |
+
{
|
| 4532 |
+
"epoch": 1.16,
|
| 4533 |
+
"mmlu_eval_accuracy": 0.48871035457552675,
|
| 4534 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
|
| 4535 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 4536 |
+
"mmlu_eval_accuracy_astronomy": 0.4375,
|
| 4537 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 4538 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
|
| 4539 |
+
"mmlu_eval_accuracy_college_biology": 0.375,
|
| 4540 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 4541 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 4542 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 4543 |
+
"mmlu_eval_accuracy_college_medicine": 0.36363636363636365,
|
| 4544 |
+
"mmlu_eval_accuracy_college_physics": 0.36363636363636365,
|
| 4545 |
+
"mmlu_eval_accuracy_computer_security": 0.36363636363636365,
|
| 4546 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
|
| 4547 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 4548 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.25,
|
| 4549 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
|
| 4550 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 4551 |
+
"mmlu_eval_accuracy_global_facts": 0.5,
|
| 4552 |
+
"mmlu_eval_accuracy_high_school_biology": 0.5,
|
| 4553 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.22727272727272727,
|
| 4554 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 4555 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 4556 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
|
| 4557 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5238095238095238,
|
| 4558 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.46511627906976744,
|
| 4559 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
|
| 4560 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
|
| 4561 |
+
"mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
|
| 4562 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
|
| 4563 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
|
| 4564 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.5909090909090909,
|
| 4565 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
|
| 4566 |
+
"mmlu_eval_accuracy_human_aging": 0.6086956521739131,
|
| 4567 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 4568 |
+
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
|
| 4569 |
+
"mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
|
| 4570 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
|
| 4571 |
+
"mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
|
| 4572 |
+
"mmlu_eval_accuracy_management": 0.5454545454545454,
|
| 4573 |
+
"mmlu_eval_accuracy_marketing": 0.76,
|
| 4574 |
+
"mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
|
| 4575 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6511627906976745,
|
| 4576 |
+
"mmlu_eval_accuracy_moral_disputes": 0.5,
|
| 4577 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.23,
|
| 4578 |
+
"mmlu_eval_accuracy_nutrition": 0.6666666666666666,
|
| 4579 |
+
"mmlu_eval_accuracy_philosophy": 0.4411764705882353,
|
| 4580 |
+
"mmlu_eval_accuracy_prehistory": 0.6571428571428571,
|
| 4581 |
+
"mmlu_eval_accuracy_professional_accounting": 0.22580645161290322,
|
| 4582 |
+
"mmlu_eval_accuracy_professional_law": 0.3588235294117647,
|
| 4583 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
|
| 4584 |
+
"mmlu_eval_accuracy_professional_psychology": 0.5072463768115942,
|
| 4585 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 4586 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 4587 |
+
"mmlu_eval_accuracy_sociology": 0.7727272727272727,
|
| 4588 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
|
| 4589 |
+
"mmlu_eval_accuracy_virology": 0.5,
|
| 4590 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 4591 |
+
"mmlu_loss": 1.1903317893609677,
|
| 4592 |
+
"step": 4800
|
| 4593 |
+
},
|
| 4594 |
+
{
|
| 4595 |
+
"epoch": 1.16,
|
| 4596 |
+
"learning_rate": 0.0002,
|
| 4597 |
+
"loss": 0.6886,
|
| 4598 |
+
"step": 4810
|
| 4599 |
+
},
|
| 4600 |
+
{
|
| 4601 |
+
"epoch": 1.17,
|
| 4602 |
+
"learning_rate": 0.0002,
|
| 4603 |
+
"loss": 0.6877,
|
| 4604 |
+
"step": 4820
|
| 4605 |
+
},
|
| 4606 |
+
{
|
| 4607 |
+
"epoch": 1.17,
|
| 4608 |
+
"learning_rate": 0.0002,
|
| 4609 |
+
"loss": 0.6717,
|
| 4610 |
+
"step": 4830
|
| 4611 |
+
},
|
| 4612 |
+
{
|
| 4613 |
+
"epoch": 1.17,
|
| 4614 |
+
"learning_rate": 0.0002,
|
| 4615 |
+
"loss": 0.6735,
|
| 4616 |
+
"step": 4840
|
| 4617 |
+
},
|
| 4618 |
+
{
|
| 4619 |
+
"epoch": 1.17,
|
| 4620 |
+
"learning_rate": 0.0002,
|
| 4621 |
+
"loss": 0.6194,
|
| 4622 |
+
"step": 4850
|
| 4623 |
+
},
|
| 4624 |
+
{
|
| 4625 |
+
"epoch": 1.18,
|
| 4626 |
+
"learning_rate": 0.0002,
|
| 4627 |
+
"loss": 0.6766,
|
| 4628 |
+
"step": 4860
|
| 4629 |
+
},
|
| 4630 |
+
{
|
| 4631 |
+
"epoch": 1.18,
|
| 4632 |
+
"learning_rate": 0.0002,
|
| 4633 |
+
"loss": 0.6295,
|
| 4634 |
+
"step": 4870
|
| 4635 |
+
},
|
| 4636 |
+
{
|
| 4637 |
+
"epoch": 1.18,
|
| 4638 |
+
"learning_rate": 0.0002,
|
| 4639 |
+
"loss": 0.7177,
|
| 4640 |
+
"step": 4880
|
| 4641 |
+
},
|
| 4642 |
+
{
|
| 4643 |
+
"epoch": 1.18,
|
| 4644 |
+
"learning_rate": 0.0002,
|
| 4645 |
+
"loss": 0.7452,
|
| 4646 |
+
"step": 4890
|
| 4647 |
+
},
|
| 4648 |
+
{
|
| 4649 |
+
"epoch": 1.19,
|
| 4650 |
+
"learning_rate": 0.0002,
|
| 4651 |
+
"loss": 0.6382,
|
| 4652 |
+
"step": 4900
|
| 4653 |
+
},
|
| 4654 |
+
{
|
| 4655 |
+
"epoch": 1.19,
|
| 4656 |
+
"learning_rate": 0.0002,
|
| 4657 |
+
"loss": 0.6749,
|
| 4658 |
+
"step": 4910
|
| 4659 |
+
},
|
| 4660 |
+
{
|
| 4661 |
+
"epoch": 1.19,
|
| 4662 |
+
"learning_rate": 0.0002,
|
| 4663 |
+
"loss": 0.6943,
|
| 4664 |
+
"step": 4920
|
| 4665 |
+
},
|
| 4666 |
+
{
|
| 4667 |
+
"epoch": 1.19,
|
| 4668 |
+
"learning_rate": 0.0002,
|
| 4669 |
+
"loss": 0.6514,
|
| 4670 |
+
"step": 4930
|
| 4671 |
+
},
|
| 4672 |
+
{
|
| 4673 |
+
"epoch": 1.2,
|
| 4674 |
+
"learning_rate": 0.0002,
|
| 4675 |
+
"loss": 0.6974,
|
| 4676 |
+
"step": 4940
|
| 4677 |
+
},
|
| 4678 |
+
{
|
| 4679 |
+
"epoch": 1.2,
|
| 4680 |
+
"learning_rate": 0.0002,
|
| 4681 |
+
"loss": 0.6723,
|
| 4682 |
+
"step": 4950
|
| 4683 |
+
},
|
| 4684 |
+
{
|
| 4685 |
+
"epoch": 1.2,
|
| 4686 |
+
"learning_rate": 0.0002,
|
| 4687 |
+
"loss": 0.6599,
|
| 4688 |
+
"step": 4960
|
| 4689 |
+
},
|
| 4690 |
+
{
|
| 4691 |
+
"epoch": 1.2,
|
| 4692 |
+
"learning_rate": 0.0002,
|
| 4693 |
+
"loss": 0.63,
|
| 4694 |
+
"step": 4970
|
| 4695 |
+
},
|
| 4696 |
+
{
|
| 4697 |
+
"epoch": 1.21,
|
| 4698 |
+
"learning_rate": 0.0002,
|
| 4699 |
+
"loss": 0.6311,
|
| 4700 |
+
"step": 4980
|
| 4701 |
+
},
|
| 4702 |
+
{
|
| 4703 |
+
"epoch": 1.21,
|
| 4704 |
+
"learning_rate": 0.0002,
|
| 4705 |
+
"loss": 0.7421,
|
| 4706 |
+
"step": 4990
|
| 4707 |
+
},
|
| 4708 |
+
{
|
| 4709 |
+
"epoch": 1.21,
|
| 4710 |
+
"learning_rate": 0.0002,
|
| 4711 |
+
"loss": 0.6848,
|
| 4712 |
+
"step": 5000
|
| 4713 |
+
},
|
| 4714 |
+
{
|
| 4715 |
+
"epoch": 1.21,
|
| 4716 |
+
"eval_loss": 0.7636929750442505,
|
| 4717 |
+
"eval_runtime": 146.981,
|
| 4718 |
+
"eval_samples_per_second": 6.804,
|
| 4719 |
+
"eval_steps_per_second": 3.402,
|
| 4720 |
+
"step": 5000
|
| 4721 |
+
},
|
| 4722 |
+
{
|
| 4723 |
+
"epoch": 1.21,
|
| 4724 |
+
"mmlu_eval_accuracy": 0.4931348567813168,
|
| 4725 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
|
| 4726 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 4727 |
+
"mmlu_eval_accuracy_astronomy": 0.375,
|
| 4728 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 4729 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
|
| 4730 |
+
"mmlu_eval_accuracy_college_biology": 0.375,
|
| 4731 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 4732 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 4733 |
+
"mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
|
| 4734 |
+
"mmlu_eval_accuracy_college_medicine": 0.36363636363636365,
|
| 4735 |
+
"mmlu_eval_accuracy_college_physics": 0.36363636363636365,
|
| 4736 |
+
"mmlu_eval_accuracy_computer_security": 0.5454545454545454,
|
| 4737 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
|
| 4738 |
+
"mmlu_eval_accuracy_econometrics": 0.08333333333333333,
|
| 4739 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.1875,
|
| 4740 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.2926829268292683,
|
| 4741 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 4742 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 4743 |
+
"mmlu_eval_accuracy_high_school_biology": 0.5,
|
| 4744 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.18181818181818182,
|
| 4745 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 4746 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 4747 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
|
| 4748 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
|
| 4749 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4883720930232558,
|
| 4750 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
|
| 4751 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
|
| 4752 |
+
"mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
|
| 4753 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.85,
|
| 4754 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
|
| 4755 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.5909090909090909,
|
| 4756 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
|
| 4757 |
+
"mmlu_eval_accuracy_human_aging": 0.6956521739130435,
|
| 4758 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 4759 |
+
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
|
| 4760 |
+
"mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
|
| 4761 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
|
| 4762 |
+
"mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
|
| 4763 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 4764 |
+
"mmlu_eval_accuracy_marketing": 0.76,
|
| 4765 |
+
"mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
|
| 4766 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
|
| 4767 |
+
"mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
|
| 4768 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.23,
|
| 4769 |
+
"mmlu_eval_accuracy_nutrition": 0.6666666666666666,
|
| 4770 |
+
"mmlu_eval_accuracy_philosophy": 0.47058823529411764,
|
| 4771 |
+
"mmlu_eval_accuracy_prehistory": 0.6285714285714286,
|
| 4772 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
|
| 4773 |
+
"mmlu_eval_accuracy_professional_law": 0.35294117647058826,
|
| 4774 |
+
"mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
|
| 4775 |
+
"mmlu_eval_accuracy_professional_psychology": 0.5072463768115942,
|
| 4776 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 4777 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 4778 |
+
"mmlu_eval_accuracy_sociology": 0.6818181818181818,
|
| 4779 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
|
| 4780 |
+
"mmlu_eval_accuracy_virology": 0.5,
|
| 4781 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 4782 |
+
"mmlu_loss": 1.2437260802981125,
|
| 4783 |
+
"step": 5000
|
| 4784 |
+
},
|
| 4785 |
+
{
|
| 4786 |
+
"epoch": 1.21,
|
| 4787 |
+
"learning_rate": 0.0002,
|
| 4788 |
+
"loss": 0.7409,
|
| 4789 |
+
"step": 5010
|
| 4790 |
+
},
|
| 4791 |
+
{
|
| 4792 |
+
"epoch": 1.21,
|
| 4793 |
+
"learning_rate": 0.0002,
|
| 4794 |
+
"loss": 0.6649,
|
| 4795 |
+
"step": 5020
|
| 4796 |
+
},
|
| 4797 |
+
{
|
| 4798 |
+
"epoch": 1.22,
|
| 4799 |
+
"learning_rate": 0.0002,
|
| 4800 |
+
"loss": 0.6908,
|
| 4801 |
+
"step": 5030
|
| 4802 |
+
},
|
| 4803 |
+
{
|
| 4804 |
+
"epoch": 1.22,
|
| 4805 |
+
"learning_rate": 0.0002,
|
| 4806 |
+
"loss": 0.7163,
|
| 4807 |
+
"step": 5040
|
| 4808 |
+
},
|
| 4809 |
+
{
|
| 4810 |
+
"epoch": 1.22,
|
| 4811 |
+
"learning_rate": 0.0002,
|
| 4812 |
+
"loss": 0.6727,
|
| 4813 |
+
"step": 5050
|
| 4814 |
+
},
|
| 4815 |
+
{
|
| 4816 |
+
"epoch": 1.22,
|
| 4817 |
+
"learning_rate": 0.0002,
|
| 4818 |
+
"loss": 0.7309,
|
| 4819 |
+
"step": 5060
|
| 4820 |
+
},
|
| 4821 |
+
{
|
| 4822 |
+
"epoch": 1.23,
|
| 4823 |
+
"learning_rate": 0.0002,
|
| 4824 |
+
"loss": 0.7316,
|
| 4825 |
+
"step": 5070
|
| 4826 |
+
},
|
| 4827 |
+
{
|
| 4828 |
+
"epoch": 1.23,
|
| 4829 |
+
"learning_rate": 0.0002,
|
| 4830 |
+
"loss": 0.6512,
|
| 4831 |
+
"step": 5080
|
| 4832 |
+
},
|
| 4833 |
+
{
|
| 4834 |
+
"epoch": 1.23,
|
| 4835 |
+
"learning_rate": 0.0002,
|
| 4836 |
+
"loss": 0.6255,
|
| 4837 |
+
"step": 5090
|
| 4838 |
+
},
|
| 4839 |
+
{
|
| 4840 |
+
"epoch": 1.23,
|
| 4841 |
+
"learning_rate": 0.0002,
|
| 4842 |
+
"loss": 0.6972,
|
| 4843 |
+
"step": 5100
|
| 4844 |
+
},
|
| 4845 |
+
{
|
| 4846 |
+
"epoch": 1.24,
|
| 4847 |
+
"learning_rate": 0.0002,
|
| 4848 |
+
"loss": 0.6413,
|
| 4849 |
+
"step": 5110
|
| 4850 |
+
},
|
| 4851 |
+
{
|
| 4852 |
+
"epoch": 1.24,
|
| 4853 |
+
"learning_rate": 0.0002,
|
| 4854 |
+
"loss": 0.7939,
|
| 4855 |
+
"step": 5120
|
| 4856 |
+
},
|
| 4857 |
+
{
|
| 4858 |
+
"epoch": 1.24,
|
| 4859 |
+
"learning_rate": 0.0002,
|
| 4860 |
+
"loss": 0.6724,
|
| 4861 |
+
"step": 5130
|
| 4862 |
+
},
|
| 4863 |
+
{
|
| 4864 |
+
"epoch": 1.24,
|
| 4865 |
+
"learning_rate": 0.0002,
|
| 4866 |
+
"loss": 0.7572,
|
| 4867 |
+
"step": 5140
|
| 4868 |
+
},
|
| 4869 |
+
{
|
| 4870 |
+
"epoch": 1.25,
|
| 4871 |
+
"learning_rate": 0.0002,
|
| 4872 |
+
"loss": 0.7052,
|
| 4873 |
+
"step": 5150
|
| 4874 |
+
},
|
| 4875 |
+
{
|
| 4876 |
+
"epoch": 1.25,
|
| 4877 |
+
"learning_rate": 0.0002,
|
| 4878 |
+
"loss": 0.7179,
|
| 4879 |
+
"step": 5160
|
| 4880 |
+
},
|
| 4881 |
+
{
|
| 4882 |
+
"epoch": 1.25,
|
| 4883 |
+
"learning_rate": 0.0002,
|
| 4884 |
+
"loss": 0.6534,
|
| 4885 |
+
"step": 5170
|
| 4886 |
+
},
|
| 4887 |
+
{
|
| 4888 |
+
"epoch": 1.25,
|
| 4889 |
+
"learning_rate": 0.0002,
|
| 4890 |
+
"loss": 0.6853,
|
| 4891 |
+
"step": 5180
|
| 4892 |
+
},
|
| 4893 |
+
{
|
| 4894 |
+
"epoch": 1.26,
|
| 4895 |
+
"learning_rate": 0.0002,
|
| 4896 |
+
"loss": 0.7083,
|
| 4897 |
+
"step": 5190
|
| 4898 |
+
},
|
| 4899 |
+
{
|
| 4900 |
+
"epoch": 1.26,
|
| 4901 |
+
"learning_rate": 0.0002,
|
| 4902 |
+
"loss": 0.7295,
|
| 4903 |
+
"step": 5200
|
| 4904 |
+
},
|
| 4905 |
+
{
|
| 4906 |
+
"epoch": 1.26,
|
| 4907 |
+
"eval_loss": 0.7666268944740295,
|
| 4908 |
+
"eval_runtime": 146.9143,
|
| 4909 |
+
"eval_samples_per_second": 6.807,
|
| 4910 |
+
"eval_steps_per_second": 3.403,
|
| 4911 |
+
"step": 5200
|
| 4912 |
+
},
|
| 4913 |
+
{
|
| 4914 |
+
"epoch": 1.26,
|
| 4915 |
+
"mmlu_eval_accuracy": 0.48383615951478276,
|
| 4916 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
|
| 4917 |
+
"mmlu_eval_accuracy_anatomy": 0.7142857142857143,
|
| 4918 |
+
"mmlu_eval_accuracy_astronomy": 0.4375,
|
| 4919 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 4920 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
|
| 4921 |
+
"mmlu_eval_accuracy_college_biology": 0.375,
|
| 4922 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 4923 |
+
"mmlu_eval_accuracy_college_computer_science": 0.09090909090909091,
|
| 4924 |
+
"mmlu_eval_accuracy_college_mathematics": 0.09090909090909091,
|
| 4925 |
+
"mmlu_eval_accuracy_college_medicine": 0.36363636363636365,
|
| 4926 |
+
"mmlu_eval_accuracy_college_physics": 0.2727272727272727,
|
| 4927 |
+
"mmlu_eval_accuracy_computer_security": 0.5454545454545454,
|
| 4928 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.34615384615384615,
|
| 4929 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 4930 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.1875,
|
| 4931 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.2682926829268293,
|
| 4932 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 4933 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 4934 |
+
"mmlu_eval_accuracy_high_school_biology": 0.46875,
|
| 4935 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.13636363636363635,
|
| 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": 0.9090909090909091,
|
| 4939 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
|
| 4940 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4883720930232558,
|
| 4941 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
|
| 4942 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
|
| 4943 |
+
"mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
|
| 4944 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.85,
|
| 4945 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
|
| 4946 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.5909090909090909,
|
| 4947 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
|
| 4948 |
+
"mmlu_eval_accuracy_human_aging": 0.6086956521739131,
|
| 4949 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 4950 |
+
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
|
| 4951 |
+
"mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
|
| 4952 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
|
| 4953 |
+
"mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
|
| 4954 |
+
"mmlu_eval_accuracy_management": 0.5454545454545454,
|
| 4955 |
+
"mmlu_eval_accuracy_marketing": 0.76,
|
| 4956 |
+
"mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
|
| 4957 |
+
"mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
|
| 4958 |
+
"mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
|
| 4959 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.27,
|
| 4960 |
+
"mmlu_eval_accuracy_nutrition": 0.6363636363636364,
|
| 4961 |
+
"mmlu_eval_accuracy_philosophy": 0.5,
|
| 4962 |
+
"mmlu_eval_accuracy_prehistory": 0.6,
|
| 4963 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
|
| 4964 |
+
"mmlu_eval_accuracy_professional_law": 0.3411764705882353,
|
| 4965 |
+
"mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
|
| 4966 |
+
"mmlu_eval_accuracy_professional_psychology": 0.5072463768115942,
|
| 4967 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 4968 |
+
"mmlu_eval_accuracy_security_studies": 0.5555555555555556,
|
| 4969 |
+
"mmlu_eval_accuracy_sociology": 0.6818181818181818,
|
| 4970 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
|
| 4971 |
+
"mmlu_eval_accuracy_virology": 0.5,
|
| 4972 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 4973 |
+
"mmlu_loss": 1.0978870882455736,
|
| 4974 |
+
"step": 5200
|
| 4975 |
+
},
|
| 4976 |
+
{
|
| 4977 |
+
"epoch": 1.26,
|
| 4978 |
+
"learning_rate": 0.0002,
|
| 4979 |
+
"loss": 0.7411,
|
| 4980 |
+
"step": 5210
|
| 4981 |
+
},
|
| 4982 |
+
{
|
| 4983 |
+
"epoch": 1.26,
|
| 4984 |
+
"learning_rate": 0.0002,
|
| 4985 |
+
"loss": 0.6982,
|
| 4986 |
+
"step": 5220
|
| 4987 |
+
},
|
| 4988 |
+
{
|
| 4989 |
+
"epoch": 1.27,
|
| 4990 |
+
"learning_rate": 0.0002,
|
| 4991 |
+
"loss": 0.6624,
|
| 4992 |
+
"step": 5230
|
| 4993 |
+
},
|
| 4994 |
+
{
|
| 4995 |
+
"epoch": 1.27,
|
| 4996 |
+
"learning_rate": 0.0002,
|
| 4997 |
+
"loss": 0.7264,
|
| 4998 |
+
"step": 5240
|
| 4999 |
+
},
|
| 5000 |
+
{
|
| 5001 |
+
"epoch": 1.27,
|
| 5002 |
+
"learning_rate": 0.0002,
|
| 5003 |
+
"loss": 0.632,
|
| 5004 |
+
"step": 5250
|
| 5005 |
+
},
|
| 5006 |
+
{
|
| 5007 |
+
"epoch": 1.27,
|
| 5008 |
+
"learning_rate": 0.0002,
|
| 5009 |
+
"loss": 0.6244,
|
| 5010 |
+
"step": 5260
|
| 5011 |
+
},
|
| 5012 |
+
{
|
| 5013 |
+
"epoch": 1.28,
|
| 5014 |
+
"learning_rate": 0.0002,
|
| 5015 |
+
"loss": 0.6633,
|
| 5016 |
+
"step": 5270
|
| 5017 |
+
},
|
| 5018 |
+
{
|
| 5019 |
+
"epoch": 1.28,
|
| 5020 |
+
"learning_rate": 0.0002,
|
| 5021 |
+
"loss": 0.6843,
|
| 5022 |
+
"step": 5280
|
| 5023 |
+
},
|
| 5024 |
+
{
|
| 5025 |
+
"epoch": 1.28,
|
| 5026 |
+
"learning_rate": 0.0002,
|
| 5027 |
+
"loss": 0.6778,
|
| 5028 |
+
"step": 5290
|
| 5029 |
+
},
|
| 5030 |
+
{
|
| 5031 |
+
"epoch": 1.28,
|
| 5032 |
+
"learning_rate": 0.0002,
|
| 5033 |
+
"loss": 0.6743,
|
| 5034 |
+
"step": 5300
|
| 5035 |
+
},
|
| 5036 |
+
{
|
| 5037 |
+
"epoch": 1.29,
|
| 5038 |
+
"learning_rate": 0.0002,
|
| 5039 |
+
"loss": 0.7596,
|
| 5040 |
+
"step": 5310
|
| 5041 |
+
},
|
| 5042 |
+
{
|
| 5043 |
+
"epoch": 1.29,
|
| 5044 |
+
"learning_rate": 0.0002,
|
| 5045 |
+
"loss": 0.7326,
|
| 5046 |
+
"step": 5320
|
| 5047 |
+
},
|
| 5048 |
+
{
|
| 5049 |
+
"epoch": 1.29,
|
| 5050 |
+
"learning_rate": 0.0002,
|
| 5051 |
+
"loss": 0.6545,
|
| 5052 |
+
"step": 5330
|
| 5053 |
+
},
|
| 5054 |
+
{
|
| 5055 |
+
"epoch": 1.29,
|
| 5056 |
+
"learning_rate": 0.0002,
|
| 5057 |
+
"loss": 0.6596,
|
| 5058 |
+
"step": 5340
|
| 5059 |
+
},
|
| 5060 |
+
{
|
| 5061 |
+
"epoch": 1.29,
|
| 5062 |
+
"learning_rate": 0.0002,
|
| 5063 |
+
"loss": 0.7142,
|
| 5064 |
+
"step": 5350
|
| 5065 |
+
},
|
| 5066 |
+
{
|
| 5067 |
+
"epoch": 1.3,
|
| 5068 |
+
"learning_rate": 0.0002,
|
| 5069 |
+
"loss": 0.7383,
|
| 5070 |
+
"step": 5360
|
| 5071 |
+
},
|
| 5072 |
+
{
|
| 5073 |
+
"epoch": 1.3,
|
| 5074 |
+
"learning_rate": 0.0002,
|
| 5075 |
+
"loss": 0.733,
|
| 5076 |
+
"step": 5370
|
| 5077 |
+
},
|
| 5078 |
+
{
|
| 5079 |
+
"epoch": 1.3,
|
| 5080 |
+
"learning_rate": 0.0002,
|
| 5081 |
+
"loss": 0.6636,
|
| 5082 |
+
"step": 5380
|
| 5083 |
+
},
|
| 5084 |
+
{
|
| 5085 |
+
"epoch": 1.3,
|
| 5086 |
+
"learning_rate": 0.0002,
|
| 5087 |
+
"loss": 0.6979,
|
| 5088 |
+
"step": 5390
|
| 5089 |
+
},
|
| 5090 |
+
{
|
| 5091 |
+
"epoch": 1.31,
|
| 5092 |
+
"learning_rate": 0.0002,
|
| 5093 |
+
"loss": 0.7667,
|
| 5094 |
+
"step": 5400
|
| 5095 |
+
},
|
| 5096 |
+
{
|
| 5097 |
+
"epoch": 1.31,
|
| 5098 |
+
"eval_loss": 0.7644040584564209,
|
| 5099 |
+
"eval_runtime": 146.9715,
|
| 5100 |
+
"eval_samples_per_second": 6.804,
|
| 5101 |
+
"eval_steps_per_second": 3.402,
|
| 5102 |
+
"step": 5400
|
| 5103 |
+
},
|
| 5104 |
+
{
|
| 5105 |
+
"epoch": 1.31,
|
| 5106 |
+
"mmlu_eval_accuracy": 0.4984692616219799,
|
| 5107 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
|
| 5108 |
+
"mmlu_eval_accuracy_anatomy": 0.7142857142857143,
|
| 5109 |
+
"mmlu_eval_accuracy_astronomy": 0.375,
|
| 5110 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 5111 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
|
| 5112 |
+
"mmlu_eval_accuracy_college_biology": 0.375,
|
| 5113 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 5114 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 5115 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 5116 |
+
"mmlu_eval_accuracy_college_medicine": 0.36363636363636365,
|
| 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": 0.1875,
|
| 5122 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.2682926829268293,
|
| 5123 |
+
"mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
|
| 5124 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 5125 |
+
"mmlu_eval_accuracy_high_school_biology": 0.5,
|
| 5126 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.22727272727272727,
|
| 5127 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 5128 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
|
| 5129 |
+
"mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
|
| 5130 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5238095238095238,
|
| 5131 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4883720930232558,
|
| 5132 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
|
| 5133 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
|
| 5134 |
+
"mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
|
| 5135 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
|
| 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": 0.7307692307692307,
|
| 5139 |
+
"mmlu_eval_accuracy_human_aging": 0.6086956521739131,
|
| 5140 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 5141 |
+
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
|
| 5142 |
+
"mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
|
| 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": 0.76,
|
| 5147 |
+
"mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
|
| 5148 |
+
"mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
|
| 5149 |
+
"mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
|
| 5150 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.25,
|
| 5151 |
+
"mmlu_eval_accuracy_nutrition": 0.6666666666666666,
|
| 5152 |
+
"mmlu_eval_accuracy_philosophy": 0.4117647058823529,
|
| 5153 |
+
"mmlu_eval_accuracy_prehistory": 0.6571428571428571,
|
| 5154 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
|
| 5155 |
+
"mmlu_eval_accuracy_professional_law": 0.35294117647058826,
|
| 5156 |
+
"mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
|
| 5157 |
+
"mmlu_eval_accuracy_professional_psychology": 0.5072463768115942,
|
| 5158 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 5159 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 5160 |
+
"mmlu_eval_accuracy_sociology": 0.6818181818181818,
|
| 5161 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 5162 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 5163 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 5164 |
+
"mmlu_loss": 1.1561562310388753,
|
| 5165 |
+
"step": 5400
|
| 5166 |
+
},
|
| 5167 |
+
{
|
| 5168 |
+
"epoch": 1.31,
|
| 5169 |
+
"learning_rate": 0.0002,
|
| 5170 |
+
"loss": 0.7687,
|
| 5171 |
+
"step": 5410
|
| 5172 |
+
},
|
| 5173 |
+
{
|
| 5174 |
+
"epoch": 1.31,
|
| 5175 |
+
"learning_rate": 0.0002,
|
| 5176 |
+
"loss": 0.6543,
|
| 5177 |
+
"step": 5420
|
| 5178 |
+
},
|
| 5179 |
+
{
|
| 5180 |
+
"epoch": 1.31,
|
| 5181 |
+
"learning_rate": 0.0002,
|
| 5182 |
+
"loss": 0.7085,
|
| 5183 |
+
"step": 5430
|
| 5184 |
+
},
|
| 5185 |
+
{
|
| 5186 |
+
"epoch": 1.32,
|
| 5187 |
+
"learning_rate": 0.0002,
|
| 5188 |
+
"loss": 0.6493,
|
| 5189 |
+
"step": 5440
|
| 5190 |
+
},
|
| 5191 |
+
{
|
| 5192 |
+
"epoch": 1.32,
|
| 5193 |
+
"learning_rate": 0.0002,
|
| 5194 |
+
"loss": 0.7247,
|
| 5195 |
+
"step": 5450
|
| 5196 |
+
},
|
| 5197 |
+
{
|
| 5198 |
+
"epoch": 1.32,
|
| 5199 |
+
"learning_rate": 0.0002,
|
| 5200 |
+
"loss": 0.7506,
|
| 5201 |
+
"step": 5460
|
| 5202 |
+
},
|
| 5203 |
+
{
|
| 5204 |
+
"epoch": 1.32,
|
| 5205 |
+
"learning_rate": 0.0002,
|
| 5206 |
+
"loss": 0.8031,
|
| 5207 |
+
"step": 5470
|
| 5208 |
+
},
|
| 5209 |
+
{
|
| 5210 |
+
"epoch": 1.33,
|
| 5211 |
+
"learning_rate": 0.0002,
|
| 5212 |
+
"loss": 0.6168,
|
| 5213 |
+
"step": 5480
|
| 5214 |
+
},
|
| 5215 |
+
{
|
| 5216 |
+
"epoch": 1.33,
|
| 5217 |
+
"learning_rate": 0.0002,
|
| 5218 |
+
"loss": 0.6377,
|
| 5219 |
+
"step": 5490
|
| 5220 |
+
},
|
| 5221 |
+
{
|
| 5222 |
+
"epoch": 1.33,
|
| 5223 |
+
"learning_rate": 0.0002,
|
| 5224 |
+
"loss": 0.6582,
|
| 5225 |
+
"step": 5500
|
| 5226 |
+
},
|
| 5227 |
+
{
|
| 5228 |
+
"epoch": 1.33,
|
| 5229 |
+
"learning_rate": 0.0002,
|
| 5230 |
+
"loss": 0.7406,
|
| 5231 |
+
"step": 5510
|
| 5232 |
+
},
|
| 5233 |
+
{
|
| 5234 |
+
"epoch": 1.34,
|
| 5235 |
+
"learning_rate": 0.0002,
|
| 5236 |
+
"loss": 0.7145,
|
| 5237 |
+
"step": 5520
|
| 5238 |
+
},
|
| 5239 |
+
{
|
| 5240 |
+
"epoch": 1.34,
|
| 5241 |
+
"learning_rate": 0.0002,
|
| 5242 |
+
"loss": 0.6468,
|
| 5243 |
+
"step": 5530
|
| 5244 |
+
},
|
| 5245 |
+
{
|
| 5246 |
+
"epoch": 1.34,
|
| 5247 |
+
"learning_rate": 0.0002,
|
| 5248 |
+
"loss": 0.6398,
|
| 5249 |
+
"step": 5540
|
| 5250 |
+
},
|
| 5251 |
+
{
|
| 5252 |
+
"epoch": 1.34,
|
| 5253 |
+
"learning_rate": 0.0002,
|
| 5254 |
+
"loss": 0.6107,
|
| 5255 |
+
"step": 5550
|
| 5256 |
+
},
|
| 5257 |
+
{
|
| 5258 |
+
"epoch": 1.35,
|
| 5259 |
+
"learning_rate": 0.0002,
|
| 5260 |
+
"loss": 0.6741,
|
| 5261 |
+
"step": 5560
|
| 5262 |
+
},
|
| 5263 |
+
{
|
| 5264 |
+
"epoch": 1.35,
|
| 5265 |
+
"learning_rate": 0.0002,
|
| 5266 |
+
"loss": 0.615,
|
| 5267 |
+
"step": 5570
|
| 5268 |
+
},
|
| 5269 |
+
{
|
| 5270 |
+
"epoch": 1.35,
|
| 5271 |
+
"learning_rate": 0.0002,
|
| 5272 |
+
"loss": 0.675,
|
| 5273 |
+
"step": 5580
|
| 5274 |
+
},
|
| 5275 |
+
{
|
| 5276 |
+
"epoch": 1.35,
|
| 5277 |
+
"learning_rate": 0.0002,
|
| 5278 |
+
"loss": 0.6051,
|
| 5279 |
+
"step": 5590
|
| 5280 |
+
},
|
| 5281 |
+
{
|
| 5282 |
+
"epoch": 1.36,
|
| 5283 |
+
"learning_rate": 0.0002,
|
| 5284 |
+
"loss": 0.7026,
|
| 5285 |
+
"step": 5600
|
| 5286 |
+
},
|
| 5287 |
+
{
|
| 5288 |
+
"epoch": 1.36,
|
| 5289 |
+
"eval_loss": 0.7640195488929749,
|
| 5290 |
+
"eval_runtime": 146.899,
|
| 5291 |
+
"eval_samples_per_second": 6.807,
|
| 5292 |
+
"eval_steps_per_second": 3.404,
|
| 5293 |
+
"step": 5600
|
| 5294 |
+
},
|
| 5295 |
+
{
|
| 5296 |
+
"epoch": 1.36,
|
| 5297 |
+
"mmlu_eval_accuracy": 0.48633169350260846,
|
| 5298 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
|
| 5299 |
+
"mmlu_eval_accuracy_anatomy": 0.7142857142857143,
|
| 5300 |
+
"mmlu_eval_accuracy_astronomy": 0.375,
|
| 5301 |
+
"mmlu_eval_accuracy_business_ethics": 0.36363636363636365,
|
| 5302 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034,
|
| 5303 |
+
"mmlu_eval_accuracy_college_biology": 0.3125,
|
| 5304 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 5305 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 5306 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 5307 |
+
"mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
|
| 5308 |
+
"mmlu_eval_accuracy_college_physics": 0.36363636363636365,
|
| 5309 |
+
"mmlu_eval_accuracy_computer_security": 0.7272727272727273,
|
| 5310 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
|
| 5311 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 5312 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.25,
|
| 5313 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.2682926829268293,
|
| 5314 |
+
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
|
| 5315 |
+
"mmlu_eval_accuracy_global_facts": 0.5,
|
| 5316 |
+
"mmlu_eval_accuracy_high_school_biology": 0.46875,
|
| 5317 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
|
| 5318 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 5319 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
|
| 5320 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
|
| 5321 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.47619047619047616,
|
| 5322 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4883720930232558,
|
| 5323 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
|
| 5324 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
|
| 5325 |
+
"mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
|
| 5326 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.85,
|
| 5327 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913,
|
| 5328 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
|
| 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": 0.9230769230769231,
|
| 5333 |
+
"mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
|
| 5334 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
|
| 5335 |
+
"mmlu_eval_accuracy_machine_learning": 0.18181818181818182,
|
| 5336 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 5337 |
+
"mmlu_eval_accuracy_marketing": 0.72,
|
| 5338 |
+
"mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
|
| 5339 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6511627906976745,
|
| 5340 |
+
"mmlu_eval_accuracy_moral_disputes": 0.5,
|
| 5341 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.25,
|
| 5342 |
+
"mmlu_eval_accuracy_nutrition": 0.6363636363636364,
|
| 5343 |
+
"mmlu_eval_accuracy_philosophy": 0.47058823529411764,
|
| 5344 |
+
"mmlu_eval_accuracy_prehistory": 0.5428571428571428,
|
| 5345 |
+
"mmlu_eval_accuracy_professional_accounting": 0.22580645161290322,
|
| 5346 |
+
"mmlu_eval_accuracy_professional_law": 0.3352941176470588,
|
| 5347 |
+
"mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
|
| 5348 |
+
"mmlu_eval_accuracy_professional_psychology": 0.463768115942029,
|
| 5349 |
+
"mmlu_eval_accuracy_public_relations": 0.5833333333333334,
|
| 5350 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 5351 |
+
"mmlu_eval_accuracy_sociology": 0.6818181818181818,
|
| 5352 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 5353 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 5354 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 5355 |
+
"mmlu_loss": 1.1303307581547035,
|
| 5356 |
+
"step": 5600
|
| 5357 |
+
},
|
| 5358 |
+
{
|
| 5359 |
+
"epoch": 1.36,
|
| 5360 |
+
"learning_rate": 0.0002,
|
| 5361 |
+
"loss": 0.7178,
|
| 5362 |
+
"step": 5610
|
| 5363 |
+
},
|
| 5364 |
+
{
|
| 5365 |
+
"epoch": 1.36,
|
| 5366 |
+
"learning_rate": 0.0002,
|
| 5367 |
+
"loss": 0.6309,
|
| 5368 |
+
"step": 5620
|
| 5369 |
+
},
|
| 5370 |
+
{
|
| 5371 |
+
"epoch": 1.36,
|
| 5372 |
+
"learning_rate": 0.0002,
|
| 5373 |
+
"loss": 0.6909,
|
| 5374 |
+
"step": 5630
|
| 5375 |
+
},
|
| 5376 |
+
{
|
| 5377 |
+
"epoch": 1.36,
|
| 5378 |
+
"learning_rate": 0.0002,
|
| 5379 |
+
"loss": 0.7302,
|
| 5380 |
+
"step": 5640
|
| 5381 |
+
},
|
| 5382 |
+
{
|
| 5383 |
+
"epoch": 1.37,
|
| 5384 |
+
"learning_rate": 0.0002,
|
| 5385 |
+
"loss": 0.6636,
|
| 5386 |
+
"step": 5650
|
| 5387 |
+
},
|
| 5388 |
+
{
|
| 5389 |
+
"epoch": 1.37,
|
| 5390 |
+
"learning_rate": 0.0002,
|
| 5391 |
+
"loss": 0.64,
|
| 5392 |
+
"step": 5660
|
| 5393 |
+
},
|
| 5394 |
+
{
|
| 5395 |
+
"epoch": 1.37,
|
| 5396 |
+
"learning_rate": 0.0002,
|
| 5397 |
+
"loss": 0.7539,
|
| 5398 |
+
"step": 5670
|
| 5399 |
+
},
|
| 5400 |
+
{
|
| 5401 |
+
"epoch": 1.37,
|
| 5402 |
+
"learning_rate": 0.0002,
|
| 5403 |
+
"loss": 0.6635,
|
| 5404 |
+
"step": 5680
|
| 5405 |
+
},
|
| 5406 |
+
{
|
| 5407 |
+
"epoch": 1.38,
|
| 5408 |
+
"learning_rate": 0.0002,
|
| 5409 |
+
"loss": 0.8093,
|
| 5410 |
+
"step": 5690
|
| 5411 |
+
},
|
| 5412 |
+
{
|
| 5413 |
+
"epoch": 1.38,
|
| 5414 |
+
"learning_rate": 0.0002,
|
| 5415 |
+
"loss": 0.7459,
|
| 5416 |
+
"step": 5700
|
| 5417 |
+
},
|
| 5418 |
+
{
|
| 5419 |
+
"epoch": 1.38,
|
| 5420 |
+
"learning_rate": 0.0002,
|
| 5421 |
+
"loss": 0.6989,
|
| 5422 |
+
"step": 5710
|
| 5423 |
+
},
|
| 5424 |
+
{
|
| 5425 |
+
"epoch": 1.38,
|
| 5426 |
+
"learning_rate": 0.0002,
|
| 5427 |
+
"loss": 0.7874,
|
| 5428 |
+
"step": 5720
|
| 5429 |
+
},
|
| 5430 |
+
{
|
| 5431 |
+
"epoch": 1.39,
|
| 5432 |
+
"learning_rate": 0.0002,
|
| 5433 |
+
"loss": 0.7504,
|
| 5434 |
+
"step": 5730
|
| 5435 |
+
},
|
| 5436 |
+
{
|
| 5437 |
+
"epoch": 1.39,
|
| 5438 |
+
"learning_rate": 0.0002,
|
| 5439 |
+
"loss": 0.7092,
|
| 5440 |
+
"step": 5740
|
| 5441 |
+
},
|
| 5442 |
+
{
|
| 5443 |
+
"epoch": 1.39,
|
| 5444 |
+
"learning_rate": 0.0002,
|
| 5445 |
+
"loss": 0.6531,
|
| 5446 |
+
"step": 5750
|
| 5447 |
+
},
|
| 5448 |
+
{
|
| 5449 |
+
"epoch": 1.39,
|
| 5450 |
+
"learning_rate": 0.0002,
|
| 5451 |
+
"loss": 0.6654,
|
| 5452 |
+
"step": 5760
|
| 5453 |
+
},
|
| 5454 |
+
{
|
| 5455 |
+
"epoch": 1.4,
|
| 5456 |
+
"learning_rate": 0.0002,
|
| 5457 |
+
"loss": 0.6015,
|
| 5458 |
+
"step": 5770
|
| 5459 |
+
},
|
| 5460 |
+
{
|
| 5461 |
+
"epoch": 1.4,
|
| 5462 |
+
"learning_rate": 0.0002,
|
| 5463 |
+
"loss": 0.6984,
|
| 5464 |
+
"step": 5780
|
| 5465 |
+
},
|
| 5466 |
+
{
|
| 5467 |
+
"epoch": 1.4,
|
| 5468 |
+
"learning_rate": 0.0002,
|
| 5469 |
+
"loss": 0.752,
|
| 5470 |
+
"step": 5790
|
| 5471 |
+
},
|
| 5472 |
+
{
|
| 5473 |
+
"epoch": 1.4,
|
| 5474 |
+
"learning_rate": 0.0002,
|
| 5475 |
+
"loss": 0.6325,
|
| 5476 |
+
"step": 5800
|
| 5477 |
+
},
|
| 5478 |
+
{
|
| 5479 |
+
"epoch": 1.4,
|
| 5480 |
+
"eval_loss": 0.7645779252052307,
|
| 5481 |
+
"eval_runtime": 146.8965,
|
| 5482 |
+
"eval_samples_per_second": 6.808,
|
| 5483 |
+
"eval_steps_per_second": 3.404,
|
| 5484 |
+
"step": 5800
|
| 5485 |
+
},
|
| 5486 |
+
{
|
| 5487 |
+
"epoch": 1.4,
|
| 5488 |
+
"mmlu_eval_accuracy": 0.48785233842472925,
|
| 5489 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
|
| 5490 |
+
"mmlu_eval_accuracy_anatomy": 0.7142857142857143,
|
| 5491 |
+
"mmlu_eval_accuracy_astronomy": 0.375,
|
| 5492 |
+
"mmlu_eval_accuracy_business_ethics": 0.36363636363636365,
|
| 5493 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
|
| 5494 |
+
"mmlu_eval_accuracy_college_biology": 0.375,
|
| 5495 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 5496 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 5497 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 5498 |
+
"mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
|
| 5499 |
+
"mmlu_eval_accuracy_college_physics": 0.36363636363636365,
|
| 5500 |
+
"mmlu_eval_accuracy_computer_security": 0.7272727272727273,
|
| 5501 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
|
| 5502 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 5503 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.1875,
|
| 5504 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.2926829268292683,
|
| 5505 |
+
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
|
| 5506 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 5507 |
+
"mmlu_eval_accuracy_high_school_biology": 0.46875,
|
| 5508 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.22727272727272727,
|
| 5509 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 5510 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.5,
|
| 5511 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
|
| 5512 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.47619047619047616,
|
| 5513 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4418604651162791,
|
| 5514 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
|
| 5515 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.5,
|
| 5516 |
+
"mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
|
| 5517 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.85,
|
| 5518 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
|
| 5519 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
|
| 5520 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
|
| 5521 |
+
"mmlu_eval_accuracy_human_aging": 0.6521739130434783,
|
| 5522 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 5523 |
+
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
|
| 5524 |
+
"mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
|
| 5525 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
|
| 5526 |
+
"mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
|
| 5527 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 5528 |
+
"mmlu_eval_accuracy_marketing": 0.8,
|
| 5529 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 5530 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
|
| 5531 |
+
"mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
|
| 5532 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.27,
|
| 5533 |
+
"mmlu_eval_accuracy_nutrition": 0.696969696969697,
|
| 5534 |
+
"mmlu_eval_accuracy_philosophy": 0.5,
|
| 5535 |
+
"mmlu_eval_accuracy_prehistory": 0.5714285714285714,
|
| 5536 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
|
| 5537 |
+
"mmlu_eval_accuracy_professional_law": 0.35294117647058826,
|
| 5538 |
+
"mmlu_eval_accuracy_professional_medicine": 0.45161290322580644,
|
| 5539 |
+
"mmlu_eval_accuracy_professional_psychology": 0.463768115942029,
|
| 5540 |
+
"mmlu_eval_accuracy_public_relations": 0.5833333333333334,
|
| 5541 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 5542 |
+
"mmlu_eval_accuracy_sociology": 0.5909090909090909,
|
| 5543 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
|
| 5544 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 5545 |
+
"mmlu_eval_accuracy_world_religions": 0.631578947368421,
|
| 5546 |
+
"mmlu_loss": 1.2788251821272678,
|
| 5547 |
+
"step": 5800
|
| 5548 |
+
},
|
| 5549 |
+
{
|
| 5550 |
+
"epoch": 1.41,
|
| 5551 |
+
"learning_rate": 0.0002,
|
| 5552 |
+
"loss": 0.7449,
|
| 5553 |
+
"step": 5810
|
| 5554 |
+
},
|
| 5555 |
+
{
|
| 5556 |
+
"epoch": 1.41,
|
| 5557 |
+
"learning_rate": 0.0002,
|
| 5558 |
+
"loss": 0.7182,
|
| 5559 |
+
"step": 5820
|
| 5560 |
+
},
|
| 5561 |
+
{
|
| 5562 |
+
"epoch": 1.41,
|
| 5563 |
+
"learning_rate": 0.0002,
|
| 5564 |
+
"loss": 0.7472,
|
| 5565 |
+
"step": 5830
|
| 5566 |
+
},
|
| 5567 |
+
{
|
| 5568 |
+
"epoch": 1.41,
|
| 5569 |
+
"learning_rate": 0.0002,
|
| 5570 |
+
"loss": 0.6144,
|
| 5571 |
+
"step": 5840
|
| 5572 |
+
},
|
| 5573 |
+
{
|
| 5574 |
+
"epoch": 1.42,
|
| 5575 |
+
"learning_rate": 0.0002,
|
| 5576 |
+
"loss": 0.655,
|
| 5577 |
+
"step": 5850
|
| 5578 |
+
},
|
| 5579 |
+
{
|
| 5580 |
+
"epoch": 1.42,
|
| 5581 |
+
"learning_rate": 0.0002,
|
| 5582 |
+
"loss": 0.6983,
|
| 5583 |
+
"step": 5860
|
| 5584 |
+
},
|
| 5585 |
+
{
|
| 5586 |
+
"epoch": 1.42,
|
| 5587 |
+
"learning_rate": 0.0002,
|
| 5588 |
+
"loss": 0.6629,
|
| 5589 |
+
"step": 5870
|
| 5590 |
+
},
|
| 5591 |
+
{
|
| 5592 |
+
"epoch": 1.42,
|
| 5593 |
+
"learning_rate": 0.0002,
|
| 5594 |
+
"loss": 0.6718,
|
| 5595 |
+
"step": 5880
|
| 5596 |
+
},
|
| 5597 |
+
{
|
| 5598 |
+
"epoch": 1.43,
|
| 5599 |
+
"learning_rate": 0.0002,
|
| 5600 |
+
"loss": 0.6714,
|
| 5601 |
+
"step": 5890
|
| 5602 |
+
},
|
| 5603 |
+
{
|
| 5604 |
+
"epoch": 1.43,
|
| 5605 |
+
"learning_rate": 0.0002,
|
| 5606 |
+
"loss": 0.6894,
|
| 5607 |
+
"step": 5900
|
| 5608 |
+
},
|
| 5609 |
+
{
|
| 5610 |
+
"epoch": 1.43,
|
| 5611 |
+
"learning_rate": 0.0002,
|
| 5612 |
+
"loss": 0.6373,
|
| 5613 |
+
"step": 5910
|
| 5614 |
+
},
|
| 5615 |
+
{
|
| 5616 |
+
"epoch": 1.43,
|
| 5617 |
+
"learning_rate": 0.0002,
|
| 5618 |
+
"loss": 0.6437,
|
| 5619 |
+
"step": 5920
|
| 5620 |
+
},
|
| 5621 |
+
{
|
| 5622 |
+
"epoch": 1.44,
|
| 5623 |
+
"learning_rate": 0.0002,
|
| 5624 |
+
"loss": 0.693,
|
| 5625 |
+
"step": 5930
|
| 5626 |
+
},
|
| 5627 |
+
{
|
| 5628 |
+
"epoch": 1.44,
|
| 5629 |
+
"learning_rate": 0.0002,
|
| 5630 |
+
"loss": 0.6233,
|
| 5631 |
+
"step": 5940
|
| 5632 |
+
},
|
| 5633 |
+
{
|
| 5634 |
+
"epoch": 1.44,
|
| 5635 |
+
"learning_rate": 0.0002,
|
| 5636 |
+
"loss": 0.6803,
|
| 5637 |
+
"step": 5950
|
| 5638 |
+
},
|
| 5639 |
+
{
|
| 5640 |
+
"epoch": 1.44,
|
| 5641 |
+
"learning_rate": 0.0002,
|
| 5642 |
+
"loss": 0.6486,
|
| 5643 |
+
"step": 5960
|
| 5644 |
+
},
|
| 5645 |
+
{
|
| 5646 |
+
"epoch": 1.44,
|
| 5647 |
+
"learning_rate": 0.0002,
|
| 5648 |
+
"loss": 0.7231,
|
| 5649 |
+
"step": 5970
|
| 5650 |
+
},
|
| 5651 |
+
{
|
| 5652 |
+
"epoch": 1.45,
|
| 5653 |
+
"learning_rate": 0.0002,
|
| 5654 |
+
"loss": 0.6419,
|
| 5655 |
+
"step": 5980
|
| 5656 |
+
},
|
| 5657 |
+
{
|
| 5658 |
+
"epoch": 1.45,
|
| 5659 |
+
"learning_rate": 0.0002,
|
| 5660 |
+
"loss": 0.7337,
|
| 5661 |
+
"step": 5990
|
| 5662 |
+
},
|
| 5663 |
+
{
|
| 5664 |
+
"epoch": 1.45,
|
| 5665 |
+
"learning_rate": 0.0002,
|
| 5666 |
+
"loss": 0.6424,
|
| 5667 |
+
"step": 6000
|
| 5668 |
+
},
|
| 5669 |
+
{
|
| 5670 |
+
"epoch": 1.45,
|
| 5671 |
+
"eval_loss": 0.7642547488212585,
|
| 5672 |
+
"eval_runtime": 146.9093,
|
| 5673 |
+
"eval_samples_per_second": 6.807,
|
| 5674 |
+
"eval_steps_per_second": 3.403,
|
| 5675 |
+
"step": 6000
|
| 5676 |
+
},
|
| 5677 |
+
{
|
| 5678 |
+
"epoch": 1.45,
|
| 5679 |
+
"mmlu_eval_accuracy": 0.5030352601745833,
|
| 5680 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.5454545454545454,
|
| 5681 |
+
"mmlu_eval_accuracy_anatomy": 0.7142857142857143,
|
| 5682 |
+
"mmlu_eval_accuracy_astronomy": 0.375,
|
| 5683 |
+
"mmlu_eval_accuracy_business_ethics": 0.36363636363636365,
|
| 5684 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
|
| 5685 |
+
"mmlu_eval_accuracy_college_biology": 0.4375,
|
| 5686 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 5687 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 5688 |
+
"mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
|
| 5689 |
+
"mmlu_eval_accuracy_college_medicine": 0.45454545454545453,
|
| 5690 |
+
"mmlu_eval_accuracy_college_physics": 0.36363636363636365,
|
| 5691 |
+
"mmlu_eval_accuracy_computer_security": 0.7272727272727273,
|
| 5692 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 5693 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 5694 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.1875,
|
| 5695 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
|
| 5696 |
+
"mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
|
| 5697 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 5698 |
+
"mmlu_eval_accuracy_high_school_biology": 0.4375,
|
| 5699 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
|
| 5700 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 5701 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
|
| 5702 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
|
| 5703 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.42857142857142855,
|
| 5704 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4883720930232558,
|
| 5705 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
|
| 5706 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
|
| 5707 |
+
"mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
|
| 5708 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.85,
|
| 5709 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913,
|
| 5710 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
|
| 5711 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
|
| 5712 |
+
"mmlu_eval_accuracy_human_aging": 0.5652173913043478,
|
| 5713 |
+
"mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
|
| 5714 |
+
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
|
| 5715 |
+
"mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
|
| 5716 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
|
| 5717 |
+
"mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
|
| 5718 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 5719 |
+
"mmlu_eval_accuracy_marketing": 0.8,
|
| 5720 |
+
"mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
|
| 5721 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
|
| 5722 |
+
"mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
|
| 5723 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.21,
|
| 5724 |
+
"mmlu_eval_accuracy_nutrition": 0.6666666666666666,
|
| 5725 |
+
"mmlu_eval_accuracy_philosophy": 0.47058823529411764,
|
| 5726 |
+
"mmlu_eval_accuracy_prehistory": 0.5428571428571428,
|
| 5727 |
+
"mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
|
| 5728 |
+
"mmlu_eval_accuracy_professional_law": 0.34705882352941175,
|
| 5729 |
+
"mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
|
| 5730 |
+
"mmlu_eval_accuracy_professional_psychology": 0.5072463768115942,
|
| 5731 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 5732 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 5733 |
+
"mmlu_eval_accuracy_sociology": 0.6818181818181818,
|
| 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": 1.0952588641534586,
|
| 5738 |
+
"step": 6000
|
| 5739 |
+
},
|
| 5740 |
+
{
|
| 5741 |
+
"epoch": 1.45,
|
| 5742 |
+
"learning_rate": 0.0002,
|
| 5743 |
+
"loss": 0.804,
|
| 5744 |
+
"step": 6010
|
| 5745 |
+
},
|
| 5746 |
+
{
|
| 5747 |
+
"epoch": 1.46,
|
| 5748 |
+
"learning_rate": 0.0002,
|
| 5749 |
+
"loss": 0.7176,
|
| 5750 |
+
"step": 6020
|
| 5751 |
+
},
|
| 5752 |
+
{
|
| 5753 |
+
"epoch": 1.46,
|
| 5754 |
+
"learning_rate": 0.0002,
|
| 5755 |
+
"loss": 0.718,
|
| 5756 |
+
"step": 6030
|
| 5757 |
+
},
|
| 5758 |
+
{
|
| 5759 |
+
"epoch": 1.46,
|
| 5760 |
+
"learning_rate": 0.0002,
|
| 5761 |
+
"loss": 0.7325,
|
| 5762 |
+
"step": 6040
|
| 5763 |
+
},
|
| 5764 |
+
{
|
| 5765 |
+
"epoch": 1.46,
|
| 5766 |
+
"learning_rate": 0.0002,
|
| 5767 |
+
"loss": 0.7166,
|
| 5768 |
+
"step": 6050
|
| 5769 |
+
},
|
| 5770 |
+
{
|
| 5771 |
+
"epoch": 1.47,
|
| 5772 |
+
"learning_rate": 0.0002,
|
| 5773 |
+
"loss": 0.7102,
|
| 5774 |
+
"step": 6060
|
| 5775 |
+
},
|
| 5776 |
+
{
|
| 5777 |
+
"epoch": 1.47,
|
| 5778 |
+
"learning_rate": 0.0002,
|
| 5779 |
+
"loss": 0.64,
|
| 5780 |
+
"step": 6070
|
| 5781 |
+
},
|
| 5782 |
+
{
|
| 5783 |
+
"epoch": 1.47,
|
| 5784 |
+
"learning_rate": 0.0002,
|
| 5785 |
+
"loss": 0.6739,
|
| 5786 |
+
"step": 6080
|
| 5787 |
+
},
|
| 5788 |
+
{
|
| 5789 |
+
"epoch": 1.47,
|
| 5790 |
+
"learning_rate": 0.0002,
|
| 5791 |
+
"loss": 0.5842,
|
| 5792 |
+
"step": 6090
|
| 5793 |
+
},
|
| 5794 |
+
{
|
| 5795 |
+
"epoch": 1.48,
|
| 5796 |
+
"learning_rate": 0.0002,
|
| 5797 |
+
"loss": 0.7177,
|
| 5798 |
+
"step": 6100
|
| 5799 |
+
},
|
| 5800 |
+
{
|
| 5801 |
+
"epoch": 1.48,
|
| 5802 |
+
"learning_rate": 0.0002,
|
| 5803 |
+
"loss": 0.6065,
|
| 5804 |
+
"step": 6110
|
| 5805 |
+
},
|
| 5806 |
+
{
|
| 5807 |
+
"epoch": 1.48,
|
| 5808 |
+
"learning_rate": 0.0002,
|
| 5809 |
+
"loss": 0.7072,
|
| 5810 |
+
"step": 6120
|
| 5811 |
+
},
|
| 5812 |
+
{
|
| 5813 |
+
"epoch": 1.48,
|
| 5814 |
+
"learning_rate": 0.0002,
|
| 5815 |
+
"loss": 0.7079,
|
| 5816 |
+
"step": 6130
|
| 5817 |
+
},
|
| 5818 |
+
{
|
| 5819 |
+
"epoch": 1.49,
|
| 5820 |
+
"learning_rate": 0.0002,
|
| 5821 |
+
"loss": 0.6881,
|
| 5822 |
+
"step": 6140
|
| 5823 |
+
},
|
| 5824 |
+
{
|
| 5825 |
+
"epoch": 1.49,
|
| 5826 |
+
"learning_rate": 0.0002,
|
| 5827 |
+
"loss": 0.6887,
|
| 5828 |
+
"step": 6150
|
| 5829 |
+
},
|
| 5830 |
+
{
|
| 5831 |
+
"epoch": 1.49,
|
| 5832 |
+
"learning_rate": 0.0002,
|
| 5833 |
+
"loss": 0.6509,
|
| 5834 |
+
"step": 6160
|
| 5835 |
+
},
|
| 5836 |
+
{
|
| 5837 |
+
"epoch": 1.49,
|
| 5838 |
+
"learning_rate": 0.0002,
|
| 5839 |
+
"loss": 0.6852,
|
| 5840 |
+
"step": 6170
|
| 5841 |
+
},
|
| 5842 |
+
{
|
| 5843 |
+
"epoch": 1.5,
|
| 5844 |
+
"learning_rate": 0.0002,
|
| 5845 |
+
"loss": 0.621,
|
| 5846 |
+
"step": 6180
|
| 5847 |
+
},
|
| 5848 |
+
{
|
| 5849 |
+
"epoch": 1.5,
|
| 5850 |
+
"learning_rate": 0.0002,
|
| 5851 |
+
"loss": 0.6887,
|
| 5852 |
+
"step": 6190
|
| 5853 |
+
},
|
| 5854 |
+
{
|
| 5855 |
+
"epoch": 1.5,
|
| 5856 |
+
"learning_rate": 0.0002,
|
| 5857 |
+
"loss": 0.655,
|
| 5858 |
+
"step": 6200
|
| 5859 |
+
},
|
| 5860 |
+
{
|
| 5861 |
+
"epoch": 1.5,
|
| 5862 |
+
"eval_loss": 0.7617577314376831,
|
| 5863 |
+
"eval_runtime": 146.9151,
|
| 5864 |
+
"eval_samples_per_second": 6.807,
|
| 5865 |
+
"eval_steps_per_second": 3.403,
|
| 5866 |
+
"step": 6200
|
| 5867 |
+
},
|
| 5868 |
+
{
|
| 5869 |
+
"epoch": 1.5,
|
| 5870 |
+
"mmlu_eval_accuracy": 0.5000325318848947,
|
| 5871 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
|
| 5872 |
+
"mmlu_eval_accuracy_anatomy": 0.7142857142857143,
|
| 5873 |
+
"mmlu_eval_accuracy_astronomy": 0.4375,
|
| 5874 |
+
"mmlu_eval_accuracy_business_ethics": 0.36363636363636365,
|
| 5875 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
|
| 5876 |
+
"mmlu_eval_accuracy_college_biology": 0.5,
|
| 5877 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 5878 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 5879 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 5880 |
+
"mmlu_eval_accuracy_college_medicine": 0.36363636363636365,
|
| 5881 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 5882 |
+
"mmlu_eval_accuracy_computer_security": 0.7272727272727273,
|
| 5883 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
|
| 5884 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 5885 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.1875,
|
| 5886 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 5887 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 5888 |
+
"mmlu_eval_accuracy_global_facts": 0.5,
|
| 5889 |
+
"mmlu_eval_accuracy_high_school_biology": 0.46875,
|
| 5890 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.18181818181818182,
|
| 5891 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 5892 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
|
| 5893 |
+
"mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
|
| 5894 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5238095238095238,
|
| 5895 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4883720930232558,
|
| 5896 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
|
| 5897 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
|
| 5898 |
+
"mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
|
| 5899 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.85,
|
| 5900 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
|
| 5901 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
|
| 5902 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
|
| 5903 |
+
"mmlu_eval_accuracy_human_aging": 0.6086956521739131,
|
| 5904 |
+
"mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
|
| 5905 |
+
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
|
| 5906 |
+
"mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
|
| 5907 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
|
| 5908 |
+
"mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
|
| 5909 |
+
"mmlu_eval_accuracy_management": 0.5454545454545454,
|
| 5910 |
+
"mmlu_eval_accuracy_marketing": 0.84,
|
| 5911 |
+
"mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
|
| 5912 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6511627906976745,
|
| 5913 |
+
"mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
|
| 5914 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.26,
|
| 5915 |
+
"mmlu_eval_accuracy_nutrition": 0.6363636363636364,
|
| 5916 |
+
"mmlu_eval_accuracy_philosophy": 0.47058823529411764,
|
| 5917 |
+
"mmlu_eval_accuracy_prehistory": 0.5714285714285714,
|
| 5918 |
+
"mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
|
| 5919 |
+
"mmlu_eval_accuracy_professional_law": 0.3411764705882353,
|
| 5920 |
+
"mmlu_eval_accuracy_professional_medicine": 0.45161290322580644,
|
| 5921 |
+
"mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
|
| 5922 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 5923 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 5924 |
+
"mmlu_eval_accuracy_sociology": 0.6363636363636364,
|
| 5925 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
|
| 5926 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 5927 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 5928 |
+
"mmlu_loss": 1.1923443862959238,
|
| 5929 |
+
"step": 6200
|
| 5930 |
}
|
| 5931 |
],
|
| 5932 |
"max_steps": 10000,
|
| 5933 |
"num_train_epochs": 3,
|
| 5934 |
+
"total_flos": 6.866328733497754e+17,
|
| 5935 |
"trial_name": null,
|
| 5936 |
"trial_params": null
|
| 5937 |
}
|
{checkpoint-4200 → checkpoint-6200}/training_args.bin
RENAMED
|
File without changes
|