GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
flavianv
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
flavianv
Model Tree
Qwen/Qwen3-4B
Input Modalities
Output Modalities
Supported Functionality
Given a natural-language outfit need (e.g. "Casual men's outfit for a weekend golf outing"), the policy operates in two stages using the same adapter, distinguished only by which system prompt/tool schema is active for a given turn:
map_outfit tool call.finalize, fix (re-search specific roles), or restart.Measured on 55-60 held-out test queries, both models at their genuinely-needed completion budgets:
| Zero-shot base | This checkpoint (interim, step 14,600) | |
|---|---|---|
| Mean reward | 0.185 | 0.571 |
| pass@1 (reward > 0.6) | 4.8% | 62.5% |
| pass@8 (reward > 0.6) | 25.5% | 83.6% |
| Mean completion length | 1118 tokens | 181 tokens |
| Roles planned (correlation with true item count) | 0.04 | 0.50 |
The gain is mostly explained by one behavioral shift: zero-shot pads nearly every outfit to ~5 roles regardless of what was actually asked; this checkpoint learned to plan almost exactly the number of items the request specifies. A held-out generalization check (10 general-knowledge QA + 10 math questions, unrelated to the outfit domain) showed no capability regression from the narrow RL fine-tuning.
Full write-up, training curves, and methodology: see the training report in the
reco-rl repo at
experiments/rl/2026-07-09_deepshopper_mapreduce_grpo_report/.
python
from peft import PeftModelfrom transformers import AutoModelForCausalLM, AutoTokenizerbase = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-4B", torch_dtype="bfloat16", device_map="cuda")model = PeftModel.from_pretrained(base, "flavianv/qwen4b-mapper-reducer-outfit-agent-grpo-lora")tokenizer = AutoTokenizer.from_pretrained("flavianv/qwen4b-mapper-reducer-outfit-agent-grpo-lora")
Use with the map_outfit/finalize_outfit-style tool schema and system prompts from
rl_map_reduce/mapper_sysprompt.md and rl_map_reduce/reducer_sysprompt.md in the reco-rl repo
to reproduce the map-reduce rollout behavior this adapter was trained under.
reco-rl repo for exact pinned versions used during training)