alt.freeweights.text / jtd-d4-34-30B
From: laion <huggingface.co/laion>Newsgroups: alt.freeweights.textSubject:

jtd-d4-34-30B

Date: 2026-11-15Commit: fcda687Signed: 2DC28CC7B26A283B ✓

Reinforcement learning model with 30.5B parameters. Safetensors weights, 61.4 GB in all. Published by laion under Apache-2.0.

30.5B params61.4GBjtd-d4-34-30Bsafetensorspermissive
FactsParams30.5BSize61.4 GBFormatssafetensorsTaskreinforcement learningLicenseApache-2.0From sourcepermissive
$ fw get laion/jtd-d4-34-30B
# manifest sha256 151dee458eae… · 61.4 GB · 65 files
$ fw verify ./jtd-d4-34-30B laion/jtd-d4-34-30B
From sourceHosted (signed)laion/jtd-d4-34-30BLast check: 2026-11-15 04:19 UTC

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

Fine-tune of Qwen3-Coder-30B-A3B-Instruct.

Works with

Runs with

Ranked by fit to this model (safetensors), then by GitHub stars.

Overview

Size
61.4 GB
Weights
61.1 GB (safetensors)
Parameters
30.5B
Formats
safetensors
Task
reinforcement learning
License
Apache-2.0 (permissive)
Files
65, each with a sha256
Publisher
laion
Added
2026-11-15
Section
alt.freeweights.text
Lineage

Original release from the publisher.

Base model: Qwen/Qwen3-Coder-30B-A3B-Instruct

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 04:19 UTCpassed

Model card

From the publisher at commit fcda687. Shown as plain text with links; images open at the publisher.

laion/jtd-d4-34-30B#

Jupiter TaskTrove DAPO campaign, arm D4 — DAPO + partial-credit arm — D1 + Harbor threshold reward shaper (1.0/0.3), filter on shaped rewards. Dataset: laion/codeforces-v2 (10,000 competitive-programming tasks; late steps trained on the TaskTrove codeforces-v3 snapshot after laion/codeforces-v2 was deleted). Base model Qwen/Qwen3-Coder-30B-A3B-Instruct; FSDP2 fully-async trainer, 16 policy GPUs (4 nodes x 4 GH200) + 4 vLLM engines (TP2); Harbor/Daytona sandboxed terminus-2 rollouts.

  • Checkpoint: step 34 of 80, selected by trailing-5 EMA of reward/avg_raw_reward over the arm's full step series (campaign ended early on platform degradation + data-quality grounds; final banked step 35).
  • Launch config: rl_config.yaml in this repo; metrics + per-link logs in training_logs/.
  • pass@8 reads 1.0 on the dynamic-sampling arms (d1/d2/d4) are post-filter by construction; use reward/avg_raw_reward series in training_logs/ for cross-arm comparison.

Training Traces#

Training-time Daytona/Harbor rollouts: open-athena/jtd-d4 — the last episode of each trial, conserved at a 25% subsample: every 4th trial dir of the arm's full trace store (33231 enumerated at upload; 167 shards, 21824 rows). Owner-instructed conservation quota for this campaign's cleanup, not the uploader default.

Files

65 files · 61.4 GBWeights 61.1 GB (safetensors). Sizes and sha256 from the signed manifest.
Show all 65 files with their sha256
FileSizesha256
.gitattributes2.14 KB6a2e7f158999cbe71cf866168898179039c58f1b1c5faf58b1d1e3d268d3ae59
README.md1.57 KBf865f68ed56c8f0d67e738498561c4fcf5f370be18fbab9aa721f82cb9b4aba3
chat_template.jinja6.21 KB5a38bfa05833266240066aedc497decc9b00cc0d3e3b8cceea98cf530196ab06
config.json1.02 KBe58feff12a5ed7fce42eb6ae6054502ba786843a515a2f6f4053adf5fb75f2bd
generation_config.json217 B2cb42f3bbca69b7dbf925354f69a3a8c5ef2184f33f8c2cd744d24aaa8e6ca69
model-00001-of-00002.safetensors50.0 GB705cc1971b21b5af07c66c08320ec205ab8c6da988431398bd967788485801e2
model-00002-of-00002.safetensors11.1 GB03ab5ac9198d063f78f9a0f85d5af86ee6a4ebb6b7692c9d1b870410b2a0d37b
model.safetensors.index.json1.70 MB8056f5e61a77bcc39322368320783ace12dfd0877cf642821620e33341d5d25e
rl_config.yaml14.2 KB46f192ada8d94c0bb1e4f11beb3179706060e68a9fb5ecae73346b6c6524b35f
tokenizer.json11.4 MBbe75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
tokenizer_config.json694 B700f87d8a901e6213c9b339ae3a0ff3f1deb2bfc39f1f66c85de3669a566fa84
training_logs/20260901_010224_metrics_jtd-d4_2_jtd-d4_1461260.csv7.00 KBb8adb045deb90c1241335ac92c62b54bb2a7c6c1d94d4be58e00c7b2f7c54423
training_logs/20260901_010224_metrics_jtd-d4_2_jtd-d4_1461261.csv7.57 KBd53e7a085d2c9085f06eb3e8654cfcdd3567f64567767ffa8ec367b7ae667772
training_logs/20260901_010224_metrics_jtd-d4_2_jtd-d4_1461262.csv7.16 KB0f512d588692167f1189ae792dd93d865777470fc8d07f6e6ea56e05e7523afb
training_logs/20260901_010224_metrics_jtd-d4_2_jtd-d4_1461263.csv7.08 KB50b8426b569174fe9920fc059f8ab11ce70cd86c9b087fa93b2e9b97cfff0ea2
training_logs/20260901_010224_metrics_jtd-d4_2_jtd-d4_1487654.csv4.08 KB2cb0273809c949a1d00ef709cb37abc0fb18fb6c7e18d7572f92aa88d17278a5
training_logs/20260901_010224_metrics_jtd-d4_2_jtd-d4_1487655.csv5.88 KB341c06103257fa0b587ea37bce6705d0f3c318c8d13678570676412a4aabdca3
training_logs/20260901_010224_metrics_jtd-d4_3_jtd-d4_1527860.csv3.96 KB6f97ab2ea2d5dd86f6eb35a24778833c466a0b8783d78dfe4af217ba805bcce8
training_logs/20260901_010224_metrics_jtd-d4_4_jtd-d4_1530846.csv3.97 KB7b2dfe5b825c41c7db0412d80c3337181cd124b6d9893253d87a83888d44ff6d
training_logs/20260901_010224_metrics_jtd-d4_jtd-d4_1459494.csv4.51 KB2a47c2ef406be4c14f8f33294c9495612e7cd10bd86215ea791c5f943320cb8d
training_logs/20260901_010224_metrics_jtd-d4_jtd-d4_1506703.csv3.96 KB289823a1ddbb334f362a6eda844a1f2f4cee4ad8cd805d5a5a62157f2e35c635
training_logs/20260901_010224_metrics_report.md31.8 KB863d5244f2a07e37bf882b06bc9b397ab89bf8f51446932763f049f7cd15f874
training_logs/20260901_010224_metrics_table.csv27.2 KBb094f91ca5d3ff70ae300c7717248d9889946fa42db1d6e456ca127de41b3682
training_logs/20260901_010224_reward_vs_steps.png195 KBeec4e42f8480ef277325c81d108b2df4049b6d27d36fce885b63c271a1c39c12
training_logs/20260901_010224_trial_results.csv9.17 MB112eebd890c2c2787949d6cea8491afa72218b8345d0d1d8e4ec58ff9780b56a
training_logs/20260901_010224_turn_count_distribution.png149 KBdcb5bc0c465409c9d64bd85d9b537f2c8ae93e282a25814b6c92161b363175b4
training_logs/20260901_010224_vllm_metrics_jtd-d4_2_jtd-d4_1461260.csv571 KB6f9501c9d2773718d6f1c52709c3be1593ce2e09316121c818859734a90a91f9
training_logs/20260901_010224_vllm_metrics_jtd-d4_2_jtd-d4_1461261.csv676 KB866564103d975adf1390109c6a51510e3538d28e5e72ce560ae3a84ae7a10f87
training_logs/20260901_010224_vllm_metrics_jtd-d4_2_jtd-d4_1461262.csv668 KB147204ca93d0d1f4e6eda8180b53461c103f33d6dec1e5eff8790deb7875ba73
training_logs/20260901_010224_vllm_metrics_jtd-d4_2_jtd-d4_1461263.csv512 KB7495c0eef730721df83bf19222e1860ac75b6fd7d5a5d21134f3413337ad2025
training_logs/20260901_010224_vllm_metrics_jtd-d4_2_jtd-d4_1487654.csv223 KBf4662099f15b9d1ea01fe8a16a82b495c7b0fbb1b3083607bb833f098be06b4b
training_logs/20260901_010224_vllm_metrics_jtd-d4_2_jtd-d4_1487655.csv659 KBcdc9a2054c3caede87143ca98458cc521d9266363ffeb51911f488c2f0887777
training_logs/20260901_010224_vllm_metrics_jtd-d4_3_jtd-d4_1527860.csv521 KB4b1164cf1430a48e78cc4ebe5e1b7dcf505a44fbed266eb606e68e06fd8323f8
training_logs/20260901_010224_vllm_metrics_jtd-d4_3_jtd-d4_1527861.csv56.8 KB4cd6ab385239718424adc7303f77aaed789937d6b3e59b8a17946d7e62013d38
training_logs/20260901_010224_vllm_metrics_jtd-d4_4_jtd-d4_1530846.csv685 KBeb03b8627a6df7e5bdef38a4d146ad9fb7c8ec6e3ddfa226a01f30bddd0d2cb8
training_logs/20260901_010224_vllm_metrics_jtd-d4_4_jtd-d4_1530847.csv673 KB1b21cb274dd916802f1d5a0de171540acb9b79df2ae43224fa4612bdb17671b1
training_logs/20260901_010224_vllm_metrics_jtd-d4_4_jtd-d4_1530848.csv48.5 KB8a5994e3b525d8a066885f31e6b2a66644340c1c8642e140f7eac10aeee52c51
training_logs/20260901_010224_vllm_metrics_jtd-d4_5_jtd-d4_1558919.csv605 KB9ecdcd85d811926729badd8e0eebd82d60f7bc65a4ee6b2c0c7bd6fc5866e076
training_logs/20260901_010224_vllm_metrics_jtd-d4_jtd-d4_1459494.csv203 KB2d7f1603e658210bd743f32531c749cb80ba4ebbe24506e1b2f116eefaa8d8ae
training_logs/20260901_010224_vllm_metrics_jtd-d4_jtd-d4_1506703.csv333 KB09cc9dddda366c2d2748103c586d0297b701420a9a6d1d1c7315b2c0f3e19e71
training_logs/20260901_010224_vllm_metrics_jtd-d4_jtd-d4_1506704.csv406 KB615d783af8d491b49e69dbd0437a77b345b8ed97747a717d44ee7d1f0dbe32f5
training_logs/20260901_010224_vllm_metrics_jtd-d4_jtd-d4_1506705.csv413 KB29999312305a161829bf4d5b95437ce3bf9af0f36077af7301c3d08e499b6e40
training_logs/20260901_010224_vllm_metrics_jtd-d4_jtd-d4_1506706.csv210 KBc584962e0490d2e503b38387a758ea89ee9f85d2ea08618ea90a24143a4036e1
training_logs/20260901_010224_vllm_metrics_table.csv9.53 MB4f3ff0e2ebbaf9fcb9f0192a68e4af3a8d81333d6928d0cc741d318f8e66d489
training_logs/jtd-d4_2_jtd-d4_1461260.out7.91 MB86ed53d3962cb9565f00ebd918873470ddef1efce29e3690eab3cbea67d6809f
training_logs/jtd-d4_2_jtd-d4_1461261.out9.36 MBcb1026f4cdaf68bfb30ddebc3695c91a9f9763e11839b9d9f2008dcbfe072624
training_logs/jtd-d4_2_jtd-d4_1461262.out9.47 MB49ba7b4391da1a5023c367117e4621c56dc10b8f75053b16f71575741fa2bdfb
training_logs/jtd-d4_2_jtd-d4_1461263.out15.1 MB7d11c8e7fa57e410a17e52693f26946a091e0fc2c834f10500b3eb2d9ba729a6
training_logs/jtd-d4_2_jtd-d4_1487654.out21.1 MB6acceb72684ec5d4879a9e5d19bcf569257448a2d7e4609cf5c29bdaec3040fa
training_logs/jtd-d4_2_jtd-d4_1487655.out9.16 MBa48a18ad2b3f9ae1e2971969dbabca5b4a939bbc869590ba7403b6bd6fe5b75f
training_logs/jtd-d4_2_jtd-d4_1487656.out4.32 KBeb6a827021d071990b8bfeba5af51a82183ef9a1e3ec613e1280d7430998c3aa
training_logs/jtd-d4_2_jtd-d4_1487657.out4.32 KBc3104d1f04e1e1233093b8ef8ca95cf4131dbef584f4117015ed4ec94f42bcf9
training_logs/jtd-d4_2_jtd-d4_1487658.out4.32 KB4a9f771492a55a6c451512b15983b2d955a82eb926356fda4f43dfc48aa0747e
training_logs/jtd-d4_3_jtd-d4_1527860.out6.82 MB751f23ca9a8b8de3f065b605640922fec4a7a882ae25b09acabb6e9ebd8d73c7
training_logs/jtd-d4_3_jtd-d4_1527861.out593 KBa726b22108157b97a3696c85228a50ef9530b53bb0ea04dddc8917072c9f8edb
training_logs/jtd-d4_4_jtd-d4_1530846.out73.5 MB5c7be194160d2e24c83381c637662d65c24a96404dbbede84281f7e544af8ac3
training_logs/jtd-d4_4_jtd-d4_1530847.out57.4 MB43741db25f10e8aa59b90b120a47977be0f05ec9175d1184648c36e81c038d10
training_logs/jtd-d4_4_jtd-d4_1530848.out3.74 MB3172016f166367a232a8461d27fd0f74b65fb5702103b414b067729ba7fe396f
training_logs/jtd-d4_5_jtd-d4_1558919.out77.2 MB0ca427bd195dc0d9dd0c616054d2b21d18d516bdfe8a2e6872861ad14413686e
training_logs/jtd-d4_jtd-d4_1459494.out2.81 MB2a197323d63f82e68ae079286e6c3fa170a7557c70f1a5d407c64041b981b38c
training_logs/jtd-d4_jtd-d4_1497348.out4.32 KBd8efd0ac83304fad8fc625628843edeeb2a95c4020bf1a5c26a1f3df455b4f85
training_logs/jtd-d4_jtd-d4_1506703.out4.28 MBe2725a05b5d130935691c6eed172a4a365fc4da9037d038ab929f5e1d5ad3925
training_logs/jtd-d4_jtd-d4_1506704.out5.06 MB40bb03b286ab77af3c3af744bc8eeb488d755a9abdd87aff525203c34dcf5529
training_logs/jtd-d4_jtd-d4_1506705.out5.82 MB96af79f1416799ba2b86d949799679bf6ce939797bcb859aca76608cc60541b3
training_logs/jtd-d4_jtd-d4_1506706.out3.25 MB5ac7d09d9e29cd295ba53b1326d4867c2fe70b95e0b1031e4c944d0b47178772
VersionsEach time the pinned upstream commit or file list changed in this repo. Newest first.
Upstream commitDateContents
fcda687636 pinned now2026-11-15 04:19 UTC65 files · 61.4 GB

Use it

$ fw get laion/jtd-d4-34-30B
# manifest sha256 151dee458eae… · 61.4 GB · 65 files
$ fw verify ./jtd-d4-34-30B laion/jtd-d4-34-30B
fw.pyChecks the signature and every sha256, tries each source in order, resumes.
$ python3 fw.py get https://freeweights.allternit.com/manifests/laion/jtd-d4-34-30B.json --dir ./jtd-d4-34-30B $ python3 fw.py verify ./jtd-d4-34-30B ./laion__jtd-d4-34-30B.json
hfThe Hugging Face CLI with one setting. Same repo id.
$ HF_ENDPOINT=https://freeweights.allternit.com hf download laion/jtd-d4-34-30B --revision fcda687636ce629438172df2a808dacf5c56c05c --local-dir ./jtd-d4-34-30B

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
f865f68ed56c8f0d67e738498561c4fcf5f370be18fbab9aa721f82cb9b4aba3
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.textSubject: jtd-d4-34-30BDate: 2026-11-15Commit: fcda687Signed: 2DC28CC7B26A283B ✓
Upstream
laion/jtd-d4-34-30B
Pinned commit
fcda687636ce629438172df2a808dacf5c56c05c
Manifest
jtd-d4-34-30B.json · .minisig
Manifest sha256
151dee458eae28f7cd23ed72b2fd86cb91a36af51ef347e665d022cdc6cd218e
Signature
minisign, Ed25519, key 2DC28CC7B26A283B
Created
2026-11-15 04:19 UTC
Generator
freeweights-manifest/0.1.0
Takedowns
None.

How to verify

$ curl -O https://freeweights.allternit.com/manifests/laion/jtd-d4-34-30B.json $ curl -O https://freeweights.allternit.com/manifests/laion/jtd-d4-34-30B.json.minisig $ minisign -Vm jtd-d4-34-30B.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-Coder-30B-A3B-Instruct · 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/jtd-d4-34-30B
Open in Freeweights Studio

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

Signed manifest
minisign, key 2DC28CC7B26A283B. jtd-d4-34-30B.json. Provenance. Takedown history: none.
From: laion <huggingface.co/laion>Subject: a2-rl-e2egit_large-50-32B
From: laion <huggingface.co/laion>Subject: a3-rl-DCAgent_exp_rpt_e2egit-large-15-8B
From: laion <huggingface.co/laion>Subject: a3-rl-DCAgent_exp_rpt_e2egit-v2-10-8B
From: laion <huggingface.co/laion>Subject: a3-rl-DCAgent_exp_rpt_pymethods2test-large-80-8B
From: laion <huggingface.co/laion>Subject: a3-rl-DCAgent_exp_rpt_unitsyn-python-large-10-8B
From: laion <huggingface.co/laion>Subject: a3-rl-DCAgent_mix_h2_language_proportional-65-8B