laion/exp-psu-swesmith-1K_glm_4-7_traces_jupiter__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/exp-psu-swesmith-1K_glm_4-7_traces_jupiter__Qwen3-8B
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-05-14 00:28:25 +08:00
commit 922bd0a571
138 changed files with 306157 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
train_results.json Normal file
View File

@@ -0,0 +1,12 @@
{
"achieved_tflops_per_gpu": 14260.634490252658,
"achieved_tflops_per_gpu_theoretical": 285996.7394798434,
"epoch": 7.0,
"mfu_percent": 1007.8186918906473,
"mfu_percent_theoretical": 20211.783708822855,
"total_flos": 3.783973798181601e+17,
"train_loss": 0.0,
"train_runtime": 0.8292,
"train_samples_per_second": 8061.816,
"train_steps_per_second": 84.417
}