Modelpublic

runtime-contracts/mistral-7b-knapsack-lora-persistent

sourceHugging Faceapache-2.0updated 16h agoView on Hugging Face

0likes

Model Card

mistral-7b-knapsack-lora-persistent

Supplementary release for the paper Evaluating Agents Across Runtime Contracts: When Mismatch Costs Efficiency or Quality (IAEval 2026, the NeurIPS 2026 Workshop on Evaluation of Interactive Agents). This is one of four LoRA adapters (Mistral-7B-v0.3 / Llama-3.1-8B base model x persistent/stateless training regime) fine-tuned on the Opaque Knapsack agentic task, extending a prior single-base-model result (see the sibling Qwen3-8B release) to a second base model family for the same paper.

—Base model: mistralai/Mistral-7B-v0.3
—Training regime: persistent (trained with a persistent Python interpreter runtime (state carries over across agent turns))
—Seed: 3407

Training configuration

Fine-tuned with Axolotl, LoRA adapter, 4-bit NF4 quantized base:

Hyperparameter	Value
lora_r	64
lora_alpha	128
lora_dropout	0.05
loratargetmodules	qproj, kproj, vproj, oproj, gateproj, upproj, down_proj
learning_rate	1e-4
lr_scheduler	cosine
optimizer	adamw_torch
epochs	3.0
microbatchsize	1
gradientaccumulationsteps	16
sequence_len	16384
sample_packing	false
seed	3407
training data	paired traces for the "persistent" regime (see paper Appendix for pairing/filtering procedure)

Mistral-7B-v0.3 ships no chattemplate -- trained with Axolotl's built-in `mistralv2v3 template; serving/eval uses a hand-written template extending it with system-role support (folded into the next user turn, since mistral_v2v3` only accepts user/assistant).

Provenance

Released alongside the paper Evaluating Agents Across Runtime Contracts: When Mismatch Costs Efficiency or Quality (IAEval 2026, the NeurIPS 2026 Workshop on Evaluation of Interactive Agents), to reproduce its reported results.

License

Apache License 2.0. The base model (Mistral-7B-v0.3) is also released under Apache 2.0.