初始化项目,由ModelHub XC社区提供模型

Model: Bykot/rugpt3small_raw_baseline
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-05-23 18:15:47 +08:00
commit 9a41c12520
11 changed files with 250686 additions and 0 deletions

View File

@@ -0,0 +1,2 @@
experiment,method,source_docs,cleaned_docs,removed_docs,removed_percent,created_at_utc
raw_baseline,raw,198000,198000,0,0.0,2026-05-05T02:43:18.823803+00:00
1 experiment method source_docs cleaned_docs removed_docs removed_percent created_at_utc
2 raw_baseline raw 198000 198000 0 0.0 2026-05-05T02:43:18.823803+00:00

View File

@@ -0,0 +1,11 @@
{
"experiment": "raw_baseline",
"method": "raw",
"base_dataset_repo": "Bykot/c4_ru_200k_split",
"source_train_docs": 198000,
"cleaned_docs": 198000,
"removed_docs": 0,
"removed_percent": 0.0,
"eval_docs": 2000,
"created_at_utc": "2026-05-05T02:43:19.833028+00:00"
}

View File

@@ -0,0 +1,2 @@
experiment,method,model_name,base_dataset_repo,train_docs,removed_docs,eval_docs,eval_loss_before,perplexity_before,eval_loss_after,perplexity_after,train_runtime,train_samples_per_second,created_at_utc
raw_baseline,raw,ai-forever/rugpt3small_based_on_gpt2,Bykot/c4_ru_200k_split,198000,0,2000,3.8492491245269775,46.95779053735424,3.0527045726776123,21.172529702158773,19125.8028,10.353,2026-05-05T08:11:53.641125+00:00
1 experiment method model_name base_dataset_repo train_docs removed_docs eval_docs eval_loss_before perplexity_before eval_loss_after perplexity_after train_runtime train_samples_per_second created_at_utc
2 raw_baseline raw ai-forever/rugpt3small_based_on_gpt2 Bykot/c4_ru_200k_split 198000 0 2000 3.8492491245269775 46.95779053735424 3.0527045726776123 21.172529702158773 19125.8028 10.353 2026-05-05T08:11:53.641125+00:00

View File

@@ -0,0 +1,16 @@
{
"experiment": "raw_baseline",
"method": "raw",
"model_name": "ai-forever/rugpt3small_based_on_gpt2",
"base_dataset_repo": "Bykot/c4_ru_200k_split",
"train_docs": 198000,
"removed_docs": 0,
"eval_docs": 2000,
"eval_loss_before": 3.8492491245269775,
"perplexity_before": 46.95779053735424,
"eval_loss_after": 3.0527045726776123,
"perplexity_after": 21.172529702158773,
"train_runtime": 19125.8028,
"train_samples_per_second": 10.353,
"created_at_utc": "2026-05-05T08:11:53.641125+00:00"
}