YAML Metadata Warning:The pipeline tag "text2text-generation" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, feature-extraction, text-generation, fill-mask, sentence-similarity, text-to-speech, text-to-audio, automatic-speech-recognition, audio-to-audio, audio-classification, audio-text-to-text, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, image-to-video, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-ranking, text-retrieval, time-series-forecasting, text-to-video, image-text-to-text, image-text-to-image, image-text-to-video, visual-question-answering, document-question-answering, zero-shot-image-classification, graph-ml, mask-generation, zero-shot-object-detection, text-to-3d, image-to-3d, image-feature-extraction, video-text-to-text, keypoint-detection, visual-document-retrieval, any-to-any, video-to-video, other

ANLP Assignment 1 - from-scratch Encoder-Decoder Transformer (C1-C5)

Ablation-study checkpoints produced by src/train.py. Implementation (see src/): attention, positional encodings, norms and the BLT modules are written from scratch, as is the byte-level BPE tokenizer (bpe_*.json caches are the full tokenizer state). Each .pt is a bare state_dict matching the constructors in src/models/.

config bit_accuracy sequence_accuracy levenshtein bleu params
C1 0.9668 0.6560 0.6820 0.9688 6048256
C2 0.5520 0.0000 31.3370 0.2849 6048256
C3 0.9120 0.3870 2.1490 0.8967 5163520
C4 0.9644 0.6300 0.7135 0.9663 6043136
C5 0.9996 0.9600 0.0455 9407236

Files

checkpoint size date
C1.pt 24.8 MB 2026-09-02 17:49
C2.pt 24.8 MB 2026-09-02 18:06
C3.pt 21.2 MB 2026-09-02 18:23
C4.pt 24.7 MB 2026-09-02 18:40
C5.pt 38.4 MB 2026-09-02 18:52
Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support