wjbmattingly/lfm2-vl-textualis

model_step_10000

Model Description

This model is a fine-tuned version of LiquidAI/LFM2-VL-450M using the brute-force-training package.

• Base Model: LiquidAI/LFM2-VL-450M
• Training Status: 🔄 In Progress
• Generated: 2025-08-18 22:55:43
• Training Steps: 10,000

Training Details

Dataset

• Dataset: wjbmattingly/catmus-edited2
• Training Examples: 94,000
• Validation Examples: 499

Training Configuration

• Max Steps: 100,000
• Batch Size: 2
• Learning Rate: 1e-05
• Gradient Accumulation: 1 steps
• Evaluation Frequency: Every 10,000 steps

Current Performance

• Training Loss: 0.294965
• Evaluation Loss: 0.455527

Pre-Training Evaluation

Initial Model Performance (before training):

• Loss: 1.270437
• Perplexity: 3.56
• Character Accuracy: 36.8%
• Word Accuracy: 17.1%

Evaluation History

All Checkpoint Evaluations

Step	Checkpoint Type	Loss	Perplexity	Char Acc	Word Acc	Improvement vs Pre
Pre	pre_training	1.2704	3.56	36.8%	17.1%	+0.0%
10,000	checkpoint	0.4555	1.58	39.0%	20.1%	+64.1%

Training Progress

Recent Training Steps (Loss Only)

Step	Training Loss	Timestamp
9,991	0.335487	2025-08-18T22:55
9,992	0.482083	2025-08-18T22:55
9,993	0.439921	2025-08-18T22:55
9,994	0.496600	2025-08-18T22:55
9,995	0.394777	2025-08-18T22:55
9,996	0.394860	2025-08-18T22:55
9,997	0.483459	2025-08-18T22:55
9,998	0.395718	2025-08-18T22:55
9,999	0.601364	2025-08-18T22:55
10,000	0.294965	2025-08-18T22:55

Training Visualizations

Training Progress and Evaluation Metrics

This chart shows the training loss progression, character accuracy, word accuracy, and perplexity over time. Red dots indicate evaluation checkpoints.

Evaluation Comparison Across All Checkpoints

Comprehensive comparison of all evaluation metrics across training checkpoints. Red=Pre-training, Blue=Checkpoints, Green=Final.

Available Visualization Files:

• training_curves.png - 4-panel view: Training loss with eval points, Character accuracy, Word accuracy, Perplexity
• evaluation_comparison.png - 4-panel comparison: Loss, Character accuracy, Word accuracy, Perplexity across all checkpoints

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer
# For vision-language models, use appropriate imports

model = AutoModelForCausalLM.from_pretrained("./model_step_10000")
tokenizer = AutoTokenizer.from_pretrained("./model_step_10000")

# Your inference code here

Training Configuration

{
  "dataset_name": "wjbmattingly/catmus-edited2",
  "model_name": "LiquidAI/LFM2-VL-450M",
  "max_steps": 100000,
  "eval_steps": 10000,
  "num_accumulation_steps": 1,
  "learning_rate": 1e-05,
  "train_batch_size": 2,
  "val_batch_size": 2,
  "train_select_start": 0,
  "train_select_end": 94000,
  "val_select_start": 94001,
  "val_select_end": 94500,
  "train_field": "Textualis",
  "val_field": "Textualis",
  "image_column": "im",
  "text_column": "text",
  "user_text": "Transcribe this medieval manuscript line.",
  "max_image_size": 200
}

Model Card Metadata

• Base Model: LiquidAI/LFM2-VL-450M
• Training Framework: brute-force-training
• Training Type: Fine-tuning
• License: Inherited from base model
• Language: Inherited from base model

---

This model card was automatically generated by brute-force-training on 2025-08-18 22:55:43