Instructions to use od2961/maxent-grpo-models with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use od2961/maxent-grpo-models with Transformers:
# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("od2961/maxent-grpo-models", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Download experiments/E112-R1/README.md from od2961/maxent-grpo-models: direct link, hf CLI and curl.
- Browser
- Download file 31.2 kB
-
https://huggingface.co/od2961/maxent-grpo-models/resolve/main/experiments/E112-R1/README.md
- Command line
-
hf download hf://od2961/maxent-grpo-models/experiments/E112-R1/README.md
-
curl -L -o README.md https://huggingface.co/od2961/maxent-grpo-models/resolve/main/experiments/E112-R1/README.md
E112-R1 · Verified-support semantic + proposal bundle
Which aspect of the update changes correctness and retained verified support?
Archive home · Study design · Coverage · Models · Comparators
Study design
Three registered model scales × five domains × five seeds. The reported two-scale contrast uses its exact common admitted intersection.
Reported exploratory bundled comparison. Semantic regularization and proposal sampling are introduced together; their individual effects are not isolated.
These are terminal exports from the registered eight-pass training horizon. The stored export-step label is 3073; the papers’ audited pass-8 evaluation endpoint is optimizer step 3072. The archive preserves that original export convention and does not imply an additional evaluated training update.
Archive coverage
| Registered new training cells | Scientific records selected | Deployable exports selected | Verified exports |
|---|---|---|---|
| 75 | 49 | 49 | 49 |
Scientific records preserve the admitted result identity. Deployable exports additionally have available model weights; verified exports have completed remote verification. Cells outside the fixed selection are not silently counted, and unverified model folders are not linked below.
| Model | Verified / selected | Seeds in selection |
|---|---|---|
| Qwen2.5-0.5B-Instruct | 25 / 25 | 43, 44, 45, 46, 47 |
| Falcon3-1B-Instruct | 24 / 24 | 55, 56, 57, 58, 59 |
Methods and collections
Verified-support semantic + proposal bundle · verified_support_discovery
Re:Dr augmented with semantic regularization and an isolated verified proposal sampler; a bundled comparison.
- Falcon3-1B-Instruct · Countdown — 4 verified exports. Exact immutable revisions are in the model tables below.
- Falcon3-1B-Instruct · Graph Coloring — 5 verified exports. Exact immutable revisions are in the model tables below.
- Falcon3-1B-Instruct · MathIR — 5 verified exports. Exact immutable revisions are in the model tables below.
- Falcon3-1B-Instruct · PantryPlan — 5 verified exports. Exact immutable revisions are in the model tables below.
- Falcon3-1B-Instruct · Python Factors — 5 verified exports. Exact immutable revisions are in the model tables below.
- Qwen2.5-0.5B-Instruct · Countdown — 5 verified exports. Exact immutable revisions are in the model tables below.
- Qwen2.5-0.5B-Instruct · Graph Coloring — 5 verified exports. Exact immutable revisions are in the model tables below.
- Qwen2.5-0.5B-Instruct · MathIR — 5 verified exports. Exact immutable revisions are in the model tables below.
- Qwen2.5-0.5B-Instruct · PantryPlan — 5 verified exports. Exact immutable revisions are in the model tables below.
- Qwen2.5-0.5B-Instruct · Python Factors — 5 verified exports. Exact immutable revisions are in the model tables below.
Verified models
Browse by domain: Countdown · Graph Coloring · Python Factors · MathIR · PantryPlan.
Countdown
9 / 9 selected exports verified.
| Model | Method | Seed | Terminal export | Immutable revision |
|---|---|---|---|---|
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 55 | step 3073 | 84e784c10b90 |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 56 | step 3073 | 240d248dbbb6 |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 57 | step 3073 | 60c7d9f38391 |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 58 | step 3073 | 27c3efc0ab3b |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 43 | step 3073 | 7871570da6a6 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 44 | step 3073 | e5904aaac84c |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 45 | step 3073 | a44155c71dbf |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 46 | step 3073 | 2fcef4be8b6a |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 47 | step 3073 | 529e54efcea6 |
Graph Coloring
10 / 10 selected exports verified.
| Model | Method | Seed | Terminal export | Immutable revision |
|---|---|---|---|---|
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 55 | step 3073 | 049be19138fb |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 56 | step 3073 | 15e2f9cc9b48 |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 57 | step 3073 | 48c020abbe0d |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 58 | step 3073 | 061deb59f113 |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 59 | step 3073 | 3faadf3ae614 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 43 | step 3073 | 930a98982984 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 44 | step 3073 | 7a4725edd91c |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 45 | step 3073 | 96b29c6269a7 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 46 | step 3073 | d8d5bf42e027 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 47 | step 3073 | 152c1903edd1 |
Python Factors
10 / 10 selected exports verified.
| Model | Method | Seed | Terminal export | Immutable revision |
|---|---|---|---|---|
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 55 | step 3073 | 7dd03c31b19e |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 56 | step 3073 | 8b80bff4895b |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 57 | step 3073 | 491cf084572d |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 58 | step 3073 | a821a38ac2b2 |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 59 | step 3073 | beaa8a21e420 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 43 | step 3073 | 8c0ac34a17a3 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 44 | step 3073 | 8d0715ce9d1d |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 45 | step 3073 | 67732dbeabb7 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 46 | step 3073 | 2a1ec42e30d5 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 47 | step 3073 | a8b381104329 |
MathIR
10 / 10 selected exports verified.
| Model | Method | Seed | Terminal export | Immutable revision |
|---|---|---|---|---|
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 55 | step 3073 | 40c31345130c |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 56 | step 3073 | 605b36c51fcc |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 57 | step 3073 | 761a7591bfa6 |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 58 | step 3073 | 28ed4e5411bd |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 59 | step 3073 | 5f83fb64df58 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 43 | step 3073 | f68546ae3373 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 44 | step 3073 | cba9a9d966c5 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 45 | step 3073 | cd9eeae075da |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 46 | step 3073 | 5158c7869ff0 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 47 | step 3073 | 39feaa03d5e6 |
PantryPlan
10 / 10 selected exports verified.
| Model | Method | Seed | Terminal export | Immutable revision |
|---|---|---|---|---|
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 55 | step 3073 | 42b58741288b |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 56 | step 3073 | 4946e1411850 |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 57 | step 3073 | f44bf8dafb24 |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 58 | step 3073 | ea860268eee4 |
| Falcon3-1B-Instruct | Verified-support semantic + proposal bundle | 59 | step 3073 | c791863b3309 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 43 | step 3073 | a3f6bb76b2c7 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 44 | step 3073 | 1f397ec84a34 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 45 | step 3073 | bb0e1a465f65 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 46 | step 3073 | a894509fbb98 |
| Qwen2.5-0.5B-Instruct | Verified-support semantic + proposal bundle | 47 | step 3073 | ba0edf50c15a |
Model-family shortcuts
Qwen2.5-0.5B-Instruct
Browse verified collections. Domain tables above provide exact model revisions.
Falcon3-1B-Instruct
Browse verified collections. Domain tables above provide exact model revisions.
Comparators and analysis scope
Use original replay comparators, with E109 repaired Python sources where required. Adaptive controllers and open-bank development branches are excluded from this reported treatment, not from every historical artifact in the broader project.
Restore and verify a model · Applicable base-model licenses. The full commit_sha, source identity and file provenance are available through the root catalog and each pinned export manifest.