31 lines
1.2 KiB
Markdown
31 lines
1.2 KiB
Markdown
---
|
||
license: apache-2.0
|
||
base_model: laion/GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink
|
||
tags:
|
||
- rl
|
||
- skyrl
|
||
- terminal-bench
|
||
- agentic
|
||
- tis
|
||
---
|
||
|
||
# explore-tis-minp-40-8B
|
||
|
||
RL (SkyRL, agentic terminal-bench / Harbor + Daytona) checkpoint from the `explore-tis` sampling-parameter ablation.
|
||
This is the **min-p sampling** arm (`explore-tis-minp`).
|
||
|
||
- **Base model:** `laion/GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink` (an 8B model)
|
||
- **Dataset:** `DCAgent/exp_rpt_pymethods2test-large`
|
||
- **Algorithm:** RLOO-n, no KL loss, `loss_reduction=seq_mean_token_sum_norm_global`, TIS on (`tis_imp_ratio_cap=2.0`)
|
||
- **Sampling:** temperature=1.0, min_p=0.05, top_k=-1, top_p=1.0
|
||
- **Selected checkpoint:** `global_step_40` (best trailing-5 EMA, α=1/3, of `reward/avg_raw_reward` over saved exports with step ≤ 78; EMA ≈ 0.4722). The 78-step cutoff was applied to exclude a step-79+ greedy/eval-pass reward artifact.
|
||
- **Max training steps:** 80
|
||
|
||
## Training Traces
|
||
|
||
Rollout traces for this run: [penfever/explore-tis-minp](https://huggingface.co/datasets/penfever/explore-tis-minp)
|
||
|
||
## Training Logs
|
||
|
||
Parsed metrics, reward-vs-steps plots, and raw console logs are under `training_logs/` in this repo.
|