GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
canyuchen
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
canyuchen
Model Tree
Qwen/Qwen2.5-1.5B-Instruct
Input Modalities
Output Modalities
Supported Functionality
rollout.n=8, train_batch 8, prompt/response 4096/512, fused kernels (triton)base_seed 42 (fully deterministic schedule)search_return_n=50), 16-env pools per client, unperturbed shared val serviceconfig/paper_accelerated/task_heterogeneity/grpo/webshop/fed_webshop_grpo_total-100_cl-per-rd-2_rd-70_ep-per-cl-3_min-goals-per-cl-100_p-hardness_success_std-1.yaml (science knobs untouched)The full run record — per-round validation dumps, training logs, curves, and the official
federated_summary.json (71 global val points + 140 per-client circles) — is published
separately as a dataset: canyuchen/fedagent-webshop-grpo-hardness-std1.
python
from transformers import AutoModelForCausalLM, AutoTokenizerrepo = "canyuchen/fedagent-webshop-grpo-hardness-std1-qwen2.5-1.5b"model = AutoModelForCausalLM.from_pretrained(repo, torch_dtype="bfloat16", device_map="auto")tok = AutoTokenizer.from_pretrained(repo)# best-val checkpoint instead of the final one:# model = AutoModelForCausalLM.from_pretrained(repo, subfolder="checkpoints/round_68", ...)
The agent expects WebShop-style observation/action prompting (as produced by the FedAgent rollout loop); it is not a general chat model, though the chat template is included.
Note on chat_template: verl's model_merger drops the tokenizer's chat_template when
exporting; it has been re-injected from the base Qwen tokenizer here, so
tok.apply_chat_template(...) works out of the box. If you re-export checkpoints from the raw
run artifacts yourself, remember to do the same.
model.safetensors.index.json included