CatalogWantedMirrorsLedgerOperating StatementDocsStudio statusCard system
alt.freeweights.qwen / Qwen3-32B-R2EGYM-256-3epochs
From: laion <huggingface.co/laion>Newsgroups: alt.freeweights.qwenSubject:

Qwen3-32B-R2EGYM-256-3epochs

Date: 2026-11-15Commit: 54ebd8aSigned: 2DC28CC7B26A283B ✓

Text generation model with 32.8B parameters. Safetensors weights, 65.5 GB in all. Published by laion under Apache-2.0.

32.8B params65.5GBQwen3-32B-R2EGYM-256-3epochssafetensorspermissive
FactsParams32.8BSize65.5 GBFormatssafetensorsTasktext generationLicenseApache-2.0From sourcepermissive
$ fw get laion/Qwen3-32B-R2EGYM-256-3epochs
# manifest sha256 bc9c4ce2c914… · 65.5 GB · 25 files
$ fw verify ./Qwen3-32B-R2EGYM-256-3epochs laion/Qwen3-32B-R2EGYM-256-3epochs
From sourceHosted (signed)laion/Qwen3-32B-R2EGYM-256-3epochsLast check: 2026-11-15 04:25 UTC
Run at homeGet help

Get help: custom fine-tune or private deploy. We review your setup before we quote.

Fine-tune of Qwen3-32B.

Works with

Run24Fine-tune29Convert15Evaluate13Scan13Apps58
OverviewModel cardFilesUse itLicenseProvenanceModel tree

Overview

Size
65.5 GB
Weights
65.5 GB (safetensors)
Parameters
32.8B
Formats
safetensors
Task
text generation
License
Apache-2.0 (permissive)
Files
25, each with a sha256
Publisher
laion
Added
2026-11-15
Section
alt.freeweights.qwen
Lineage

Original release from the publisher.

Base model: Qwen/Qwen3-32B

Where these bytes come from
Publisher at the pinned commit→You

Sources are tried in order. A file counts only when its size and sha256 match the manifest. Sources are checked when you ask for them, not on a timer.

Sources
BadgeHostLast checkResult
From sourcehuggingface.co2026-11-15 04:25 UTCpassed

Model card

From the publisher at commit 54ebd8a. Shown as plain text with links; images open at the publisher.

Qwen3-32B-R2EGYM-256-3epochs#

This model is a reinforcement learning fine-tuned version of Qwen/Qwen3-32B, trained using the SkyRL framework with fully asynchronous PPO on coding and reasoning tasks from the R2EGYM benchmark.

Training Details#

Framework#
  • Training Framework: SkyRL (fully async PPO)
  • Parallelism Strategy: FSDP2 with CPU offload
  • Agent: Terminus-2 (terminal-based coding agent with thinking enabled)
Dataset#
Hyperparameters#
ParameterValue
Epochs3
Total steps12 (4 steps/epoch)
Learning rate1e-5
Weight decay0.0
Train batch size64
Micro train batch size per GPU1
Advantage estimatorrloo_n
KL lossdisabled
Samples per prompt8
Max prompt length2,048
Max generate length30,720
RoPE scalingyarn (factor=4.0, original_max_position_embeddings=32,768)
Infrastructure#
ComponentConfiguration
Policy nodes4 nodes x 4 GPUs
Reference model nodes4 nodes x 4 GPUs
Inference engines26 (tensor parallelism = 2)
Parallel generation workers96
Concurrent sandbox trials96
Total training nodes17
Training Notes#
  • Training was resumed from a step-9 checkpoint
  • The model uses Terminus-2, a terminal-based coding agent that interacts with sandboxed Docker environments to solve programming tasks
  • Thinking mode was enabled during training (--enable_thinking)

Usage#

This model can be used as a drop-in replacement for Qwen3-32B with improved coding and reasoning capabilities.

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("laion/Qwen3-32B-R2EGYM-256-3epochs")
tokenizer = AutoTokenizer.from_pretrained("laion/Qwen3-32B-R2EGYM-256-3epochs")

Files

25 files · 65.5 GBWeights 65.5 GB (safetensors). Sizes and sha256 from the signed manifest.
Show all 25 files with their sha256
FileSizesha256
.gitattributes1.57 KB34448b82c17d60fec9b65b1f093c115ddbaadc04beb1b0140b6bfed2e012a930
README.md2.50 KB9c926e790851288b361a786767d484f44b01f7e39e4acb67a59ff20f788b95cb
added_tokens.json707 Bc0284b582e14987fbd3d5a2cb2bd139084371ed9acbae488829a1c900833c680
chat_template.jinja4.17 KBa55ee1b1660128b7098723e0abcd92caa0788061051c62d51cbe87d9cf1974d8
config.json2.27 KBe76b96d0b312e0e8752014833e92605c730fdf8c2ac87b8a4d8bf9f6fed458f7
generation_config.json214 B19d9395705449004ddebeeaf62841338a48cc8df85cddb1ca0b2c60396d95c96
merges.txt1.67 MB8831e4f1a044471340f7c0a83d7bd71306a5b867e95fd870f74d0c5308a904d5
model-00001-of-00013.safetensors5.32 GBb05e78bc8fdbaef23321fd4a3a72fbeb2690c190d9834d77c2944ce4ed8c52d1
model-00002-of-00013.safetensors5.14 GB9e3db2c0703ce0d6c15408eb4d647ca270236568f84b2e9552f51317bc8ef51a
model-00003-of-00013.safetensors5.33 GBd619cf8892ecb267344250e2ae7a3ecb2a0a7a65afcbbbecd9a8e0632a1eecd3
model-00004-of-00013.safetensors5.14 GB1ad3aecfe86e4cb91b6fda4686904b33b7239dc6edc8fe2de03a0f027fc4ceab
model-00005-of-00013.safetensors5.14 GB3197b94a9336c33e68e3a3823c24df601ace0fd2c198064af83f557259f8d558
model-00006-of-00013.safetensors5.33 GB894d83ff71ff030872d3821e7677fb29ae8552b683a30bb1c46170b62a1577ff
model-00007-of-00013.safetensors5.14 GB71c68e94c08e75b6e1177e2eb30cf76754c30886da9275ecfa3d6482552f8d24
model-00008-of-00013.safetensors5.14 GB4b310462fda3af340560cd87f0726d672be492e101c4176b4ae7dcbe74de5d20
model-00009-of-00013.safetensors5.33 GB0f6f148e3d037f9062ba006c179600272ab7b897e7a3ae75a8abed12e24c2b75
model-00010-of-00013.safetensors5.14 GB5413ddc56c8df50d962613447a582ab5f80983d140f6e7d04471177a82d7397d
model-00011-of-00013.safetensors5.14 GBf13c4fe21d4f4c7e8d76d486a33039bc51149f2d5c5131fd02e8dba8e1c56c92
model-00012-of-00013.safetensors5.33 GBd275ef446b1ef573f3cb11fe82077add4f7c54eaa63ed6149e81056cce97261e
model-00013-of-00013.safetensors2.93 GB51528a46e722205224082ab740573f7f78c92e9ce658cdf9949e83395b9850b3
model.safetensors.index.json58.3 KB046cb77526bba0b2b7ebdf21aa2a5796653675a98a2b268b79271c760b030c7b
special_tokens_map.json613 B76862e765266b85aa9459767e33cbaf13970f327a0e88d1c65846c2ddd3a1ecd
tokenizer.json11.4 MBaeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
tokenizer_config.json5.40 KB443bfa629eb16387a12edbf92a76f6a6f10b2af3b53d87ba1550adfcf45f7fa0
vocab.json2.78 MBca10d7e9fb3ed18575dd1e277a2579c16d108e32f27439684afa0e10b1440910
VersionsEach time the pinned upstream commit or file list changed in this repo. Newest first.
Upstream commitDateContents
54ebd8ab83 pinned now2026-11-15 04:25 UTC25 files · 65.5 GB

Use it

$ fw get laion/Qwen3-32B-R2EGYM-256-3epochs
# manifest sha256 bc9c4ce2c914… · 65.5 GB · 25 files
$ fw verify ./Qwen3-32B-R2EGYM-256-3epochs laion/Qwen3-32B-R2EGYM-256-3epochs
fw.pyChecks the signature and every sha256, tries each source in order, resumes.
$ python3 fw.py get https://freeweights.allternit.com/manifests/laion/Qwen3-32B-R2EGYM-256-3epochs.json --dir ./Qwen3-32B-R2EGYM-256-3epochs $ python3 fw.py verify ./Qwen3-32B-R2EGYM-256-3epochs ./laion__Qwen3-32B-R2EGYM-256-3epochs.json
hfThe Hugging Face CLI with one setting. Same repo id.
$ HF_ENDPOINT=https://freeweights.allternit.com hf download laion/Qwen3-32B-R2EGYM-256-3epochs --revision 54ebd8ab834f75219c43295767876755993ef4ca --local-dir ./Qwen3-32B-R2EGYM-256-3epochs
transformersLoad the weights in Python.
import os os.environ["HF_ENDPOINT"] = "https://freeweights.allternit.com" from transformers import AutoModelForCausalLM, AutoTokenizer tok = AutoTokenizer.from_pretrained("laion/Qwen3-32B-R2EGYM-256-3epochs", revision="54ebd8ab834f75219c43295767876755993ef4ca") model = AutoModelForCausalLM.from_pretrained("laion/Qwen3-32B-R2EGYM-256-3epochs", revision="54ebd8ab834f75219c43295767876755993ef4ca", device_map="auto")
vLLMServe an OpenAI-compatible API.
$ HF_ENDPOINT=https://freeweights.allternit.com vllm serve laion/Qwen3-32B-R2EGYM-256-3epochs --revision 54ebd8ab834f75219c43295767876755993ef4ca
Run at home

License

Name
Apache License 2.0
Id
Apache-2.0
Class
Permissive
Redistribution
yes
Listing
Hosted (signed)
Decided
License as declared by the publisher. No person read it. 2026-11-15
Text in
README.md
Text sha256
9c926e790851288b361a786767d484f44b01f7e39e4acb67a59ff20f788b95cb
Terms gate
No extra terms beyond the license.

Read the license text

Licenses are checked against the license table. A custom license passes only after a person reads it.

Provenance

From: laion <huggingface.co/laion>Newsgroups: alt.freeweights.qwenSubject: Qwen3-32B-R2EGYM-256-3epochsDate: 2026-11-15Commit: 54ebd8aSigned: 2DC28CC7B26A283B ✓
Upstream
laion/Qwen3-32B-R2EGYM-256-3epochs
Pinned commit
54ebd8ab834f75219c43295767876755993ef4ca
Manifest
Qwen3-32B-R2EGYM-256-3epochs.json · .minisig
Manifest sha256
bc9c4ce2c9141937aabe51f1e2f4fe3514419cb32dacef526c251c87916e4e52
Signature
minisign, Ed25519, key 2DC28CC7B26A283B
Created
2026-11-15 04:25 UTC
Generator
freeweights-manifest/0.1.0
Takedowns
None.

How to verify

$ curl -O https://freeweights.allternit.com/manifests/laion/Qwen3-32B-R2EGYM-256-3epochs.json $ curl -O https://freeweights.allternit.com/manifests/laion/Qwen3-32B-R2EGYM-256-3epochs.json.minisig $ minisign -Vm Qwen3-32B-R2EGYM-256-3epochs.json -P RWQ7KGqyx4zCLXtUFOqvfv8L3g+8fpKJ1FsuxPyOYMF+4KufbUxKq1Oq $ shasum -a 256 <file> # compare with the Files tab

A manifest whose signature fails must be rejected. A file whose size or sha256 does not match must be discarded. Verify a download

Model tree

Parent

Qwen/Qwen3-32B · this model is a finetune of it

No model in the catalog names this one as its parent.

Get it
Download
$ HF_ENDPOINT=https://freeweights.allternit.com hf download laion/Qwen3-32B-R2EGYM-256-3epochs
Open in Freeweights Studio

Pull, verify and run it in the Allternit app on Desktop, web or phone.

Signed manifest
minisign, key 2DC28CC7B26A283B. Qwen3-32B-R2EGYM-256-3epochs.json. Provenance. Takedown history: none.
Source
huggingface.co/laion/Qwen3-32B-R2EGYM-256-3epochs

Related models

All of alt.freeweights.qwen →
From: BAAI <huggingface.co/BAAI>Subject: Aquila-VL-2B-llava-qwen
From: Perciqa <huggingface.co/Perciqa>Subject: Aurora-Code-1
From: BAAI <huggingface.co/BAAI>Subject: bge-reasoner-embed-qwen3-8b-0923
From: amazon <huggingface.co/amazon>Subject: BMOJOF-primed-HQwen3-8B-Instruct
From: nbeerbower <huggingface.co/nbeerbower>Subject: Bubba-Qwen3.6-27B-LoRA
From: bartowski <huggingface.co/bartowski>Subject: bytkim_Qwen3.8-27B-pi-GGUF