Chess model submitted to the LLM Course Chess Challenge.

Submission Info

  • Submitted by: janisaiad
  • Parameters: 43,104
  • Organization: LLM-course

Model Details

  • Architecture: Tiny Recursive Model (TRM) - looping recurrent transformer (cycle-shared weights)
  • Vocab size: 148
  • Embedding dim: 48
  • Layers: 1
  • Heads: 2
  • Cycles: 8

TRM note: this is a looping TRM model — at inference/training time we run the same transformer stack for 8 recurrent refinement cycle(s) (weights are shared across cycles), which increases compute/reasoning depth without increasing parameter count.

Training Information

Training Metrics:

  • Best Eval Loss: 1.21398
  • Final Train Loss: 1.25820
  • Total Epochs: 6.00
  • Total Steps: 83,340

Training Loss Curves:

Note: Install matplotlib to generate loss curve plots

Training Loss History (Summary):

Step Epoch Train Loss Eval Loss Learning Rate
100 0.01 4.9583 - 5.94e-05
6,300 0.76 1.4557 - 9.43e-04
12,500 1.50 1.3543 - 8.67e-04
18,700 2.24 1.3177 - 7.91e-04
24,900 2.99 1.2974 - 7.16e-04
31,100 3.73 1.2870 - 6.40e-04
37,300 4.48 1.2731 - 5.64e-04
43,500 5.22 1.2620 - 4.88e-04
49,700 5.96 1.2557 - 4.12e-04
50,000 6.00 - 1.21398 -
Downloads last month
2
Safetensors
Model size
43.1k params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support