vital-ar/physics-solver-e4-adversarial
Viewer β’ Updated β’ 1.52k β’ 106
How to use vital-ar/adversarial-physics-solver-e4-adversarial-lora with PEFT:
from peft import PeftModel
from transformers import AutoModelForCausalLM
base_model = AutoModelForCausalLM.from_pretrained("CohereLabs/tiny-aya-global")
model = PeftModel.from_pretrained(base_model, "vital-ar/adversarial-physics-solver-e4-adversarial-lora")Part of Adversarial Physics SolverβVerifier (Track 1: Data-Scarce University Physics). Multi-turn hint-conditioned Solver (CohereLabs/tiny-aya-global, Model A) designed to solve Turn-0 physics problems and revise derivations when given self-contained [[CHALLENGE]] directives from the E4b Verifier without repeating mistakes. Extends E3 Turn-0 derivations with harvested valid tiny-aya-global attempts, aya-expanse-8b Alternative Solution 2s, and Turn-1/Turn-2 corrective hint trajectories.
vital-ar/physics-solver-e4-adversarial (1444 training examples; strict < 0.80 MiniLM separation vs. Clean held-out splits)763, Turn 1 (1 hint): 340, Turn 2 (2 hints): 341 | Sources: literature_curated: 1418, physgym: 26ELECTROMAGNETISM 300, MECHANICS 770, MODERN_PHYSICS 128, OPTICS 72, RELATIVITY 108, THERMODYNAMICS 66. Data filtering/generation used CohereForAI/aya-expanse-8b exclusively; training & failure harvesting used CohereLabs/tiny-aya-global exclusively.2.0e-4 Solvers / 1.5e-4 Verifier, LoRA 16/32, max_epochs=10, raw unnormalized steps).AlignTune aligntune==0.1.32): Base CohereLabs/tiny-aya-global (3.35B FP16), LoRA r=16 / alpha=32, LR 0.0002, epochs 4.0/10, steps 364, final train loss 0.4349, selected checkpoint final, completion_only_loss=true, seed=42.cosine < 0.80 Splits)
| Clean Split | Overall Acc | Textbook Acc | Sentinel % | Degenerate % | AuditKIT SHA-256 |
|---|---|---|---|---|---|
validation (n=38) |
0.00% (0/38) | N/A (PhysGym only) | 63.16% | 36.84% | ae55f7a7347cf5f7 |
test (n=37) |
0.00% (0/37) | N/A (PhysGym only) | 59.46% | 40.54% | fc8d4ca448a6440c |
YimengChen/PhysGym (38 problems) stays 0/38 across all runs (see main repo for root-cause analysis and Table 1 pre-filter comparison).AlignTune Raw Prompt Format)
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel
base_id = "CohereLabs/tiny-aya-global"
adapter_id = "vital-ar/adversarial-physics-solver-e4-adversarial-lora"
tok = AutoTokenizer.from_pretrained(base_id)
base = AutoModelForCausalLM.from_pretrained(base_id, torch_dtype=torch.float16, device_map="auto")
model = PeftModel.from_pretrained(base, adapter_id).eval()
prompt = (
"Solve the following physics problem step by step (in at most 10 logical steps). "
"State the governing physical equations, perform algebraic substitutions clearly, include "
"SI units, and end with the final answer on a new line prefixed with ANSWER:.\n\n"
"PROBLEM:\nA 2 kg block starts from rest on a 5 m frictionless ramp. Find its speed "
"at the bottom.\n\n"
)
inputs = tok(prompt, return_tensors="pt").to(model.device)
out = model.generate(**inputs, max_new_tokens=1024, do_sample=False)
print(tok.decode(out[0][inputs.input_ids.shape[1]:], skip_special_tokens=True))
@vital-ar | Repo: GitHub | Collections: Models (15) Β· Datasets & Traces