laion/100k_warmup0.05__Qwen3-8B
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

初始化项目，由ModelHub XC社区提供模型

Browse Source 
Model: laion/100k_warmup0.05__Qwen3-8B
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-05-26 20:24:21 +08:00
commit ab9ede060f
23 changed files with 163400 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
all_results.json Normal file
View File

@@ -0,0 +1,16 @@
{
"achieved_tflops_per_gpu": 91.47451401577693,
"achieved_tflops_per_gpu_theoretical": 13518477.768859137,
"epoch": 7.0,
"loss_nan_ranks": 0,
"loss_rank_avg": 0.2246013879776001,
"mfu_percent": 6.464629965779288,
"mfu_percent_theoretical": 955369.453629621,
"total_flos": 1.326951254196224e+16,
"train_loss": 0.0,
"train_runtime": 1.1333,
"train_samples_per_second": 510432.869,
"train_steps_per_second": 3990.218,
"valid_targets_mean": 2756.4,
"valid_targets_min": 182
}