Upload folder using huggingface_hub

README.md

@@ -0,0 +1,85 @@
+---
+language: kik
+tags:
+  - automatic-speech-recognition
+  - w2v-bert-2.0
+  - kikuyu
+  - low-resource
+  - adapter
+  - peft
+license: apache-2.0
+base_model: facebook/w2v-bert-2.0
+datasets:
+  - mutisya/Kikuyu_asr_v24_23_1-filtered
+metrics:
+  - wer
+model-index:
+  - name: w2v-bert-hybrid-v3-kikuyu-asr
+    results:
+      - task:
+          type: automatic-speech-recognition
+          name: Speech Recognition
+        dataset:
+          name: Kikuyu ASR
+          type: mutisya/Kikuyu_asr_v24_23_1-filtered
+        metrics:
+          - type: wer
+            value: 20.30
+            name: WER
+---
+
+# W2V-BERT 2.0 Hybrid V3 Kikuyu ASR
+
+This model is a fine-tuned version of [facebook/w2v-bert-2.0](https://huggingface.co/facebook/w2v-bert-2.0) for Kikuyu (Gĩkũyũ) automatic speech recognition.
+
+## Model Description
+
+This model uses a **Hybrid V3 architecture** that combines:
+- **24 MMS-style bottleneck adapters** (64-dim) in each transformer layer
+- **Single-layer transformer decoder** with pre+post normalization
+- **Gated residual connections** for stable training
+
+### Architecture Details
+
+- **Base Model**: facebook/w2v-bert-2.0 (580M parameters)
+- **Trainable Parameters**: 11,660,835 (1.97% of total)
+- **Adapter Dimension**: 64
+- **Decoder Hidden Size**: 1024 (matches W2V-BERT)
+- **Decoder FFN Size**: 2048
+
+## Training Details
+
+- **Training Samples**: 5,000
+- **Epochs**: 20
+- **Learning Rate**: 0.0003
+- **Batch Size**: 4 (effective: 16 with gradient accumulation)
+- **Warmup Steps**: 500
+- **Optimizer**: AdamW with cosine LR schedule
+
+## Performance
+
+| Metric | Value |
+|--------|-------|
+| **Word Error Rate (WER)** | **20.30%** |
+| Eval Loss | 0.2371 |
+| Train Loss | 0.3413 |
+
+## Usage
+
+
+
+## Limitations
+
+- Trained specifically for Kikuyu language
+- Best performance on clean, clear audio
+- May struggle with heavy background noise or very fast speech
+
+## Citation
+
+If you use this model, please cite:
+
+
+
+## License
+
+Apache 2.0

added_tokens.json

@@ -0,0 +1,4 @@
+{
+  "</s>": 34,
+  "<s>": 33
+}

config.json

@@ -0,0 +1,81 @@
+{
+  "activation_dropout": 0.0,
+  "adapter_act": "relu",
+  "adapter_kernel_size": 3,
+  "adapter_stride": 2,
+  "add_adapter": false,
+  "apply_spec_augment": false,
+  "architectures": [
+    "Wav2Vec2BertForCTCHybridV3"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 33,
+  "classifier_proj_size": 768,
+  "codevector_dim": 768,
+  "conformer_conv_dropout": 0.1,
+  "contrastive_logits_temperature": 0.1,
+  "conv_depthwise_kernel_size": 31,
+  "ctc_loss_reduction": "mean",
+  "ctc_zero_infinity": true,
+  "diversity_loss_weight": 0.1,
+  "dtype": "float32",
+  "eos_token_id": 34,
+  "feat_proj_dropout": 0.0,
+  "feat_quantizer_dropout": 0.0,
+  "feature_projection_input_dim": 160,
+  "final_dropout": 0.1,
+  "hidden_act": "swish",
+  "hidden_dropout": 0.0,
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "layer_norm_eps": 1e-05,
+  "layerdrop": 0.1,
+  "left_max_position_embeddings": 64,
+  "mask_feature_length": 10,
+  "mask_feature_min_masks": 0,
+  "mask_feature_prob": 0.0,
+  "mask_time_length": 10,
+  "mask_time_min_masks": 2,
+  "mask_time_prob": 0.05,
+  "max_source_positions": 5000,
+  "model_type": "wav2vec2-bert",
+  "num_adapter_layers": 1,
+  "num_attention_heads": 16,
+  "num_codevector_groups": 2,
+  "num_codevectors_per_group": 320,
+  "num_hidden_layers": 24,
+  "num_negatives": 100,
+  "output_hidden_size": 1024,
+  "pad_token_id": 32,
+  "position_embeddings_type": "relative_key",
+  "proj_codevector_dim": 768,
+  "right_max_position_embeddings": 8,
+  "rotary_embedding_base": 10000,
+  "tdnn_dilation": [
+    1,
+    2,
+    3,
+    1,
+    1
+  ],
+  "tdnn_dim": [
+    512,
+    512,
+    512,
+    512,
+    1500
+  ],
+  "tdnn_kernel": [
+    5,
+    3,
+    3,
+    1,
+    1
+  ],
+  "transformers_version": "4.57.1",
+  "use_intermediate_ffn_before_adapter": false,
+  "use_weighted_layer_sum": false,
+  "vocab_size": 33,
+  "xvector_output_dim": 512
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a7452bb482f2ceecd48116e2f1d2d0a07bd09d3b8add8d6119f5e6325c309da
+size 2368733780

preprocessor_config.json

@@ -0,0 +1,11 @@
+{
+  "feature_extractor_type": "SeamlessM4TFeatureExtractor",
+  "feature_size": 80,
+  "num_mel_bins": 80,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "processor_class": "Wav2Vec2BertProcessor",
+  "return_attention_mask": true,
+  "sampling_rate": 16000,
+  "stride": 2
+}

runs/events.out.tfevents.1764214770.Denali.173540.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63d91b1f71c978fad90540e2dccd41e39ed7c030ddb76698be9636c0227dab10
+size 25148

runs/events.out.tfevents.1764216368.Denali.173540.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c763e9745308dec08aff1766c80906e1c9e3031150095774e1b4b7faab4ebcfa
+size 406

runs/events.out.tfevents.1764223754.Denali.201540.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5cf8df81702cefe6ebefeef57c5c88646fb54b4ab9c2d659041a7e05e248a606
+size 99148

runs/events.out.tfevents.1764244637.Denali.201540.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c886d396d61d5f3a8c6f3f2d39f171e810f36fa9f310cdf98a7cb182da8c0b7
+size 406

special_tokens_map.json

@@ -0,0 +1,30 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "[PAD]",
+    "lstrip": true,
+    "normalized": false,
+    "rstrip": true,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "[UNK]",
+    "lstrip": true,
+    "normalized": false,
+    "rstrip": true,
+    "single_word": false
+  }
+}

tokenizer_config.json

@@ -0,0 +1,49 @@
+{
+  "added_tokens_decoder": {
+    "31": {
+      "content": "[UNK]",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": false
+    },
+    "32": {
+      "content": "[PAD]",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": false
+    },
+    "33": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "34": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "do_lower_case": false,
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "[PAD]",
+  "processor_class": "Wav2Vec2BertProcessor",
+  "replace_word_delimiter_char": " ",
+  "target_lang": null,
+  "tokenizer_class": "Wav2Vec2CTCTokenizer",
+  "unk_token": "[UNK]",
+  "word_delimiter_token": "|"
+}

vocab.json

@@ -0,0 +1,35 @@
+{
+  "'": 1,
+  "-": 2,
+  "[PAD]": 32,
+  "[UNK]": 31,
+  "a": 3,
+  "b": 4,
+  "c": 5,
+  "d": 6,
+  "e": 7,
+  "f": 8,
+  "g": 9,
+  "h": 10,
+  "i": 11,
+  "j": 12,
+  "k": 13,
+  "l": 14,
+  "m": 15,
+  "n": 16,
+  "o": 17,
+  "p": 18,
+  "q": 19,
+  "r": 20,
+  "s": 21,
+  "t": 22,
+  "u": 23,
+  "v": 24,
+  "w": 25,
+  "x": 26,
+  "y": 27,
+  "z": 28,
+  "|": 0,
+  "ĩ": 29,
+  "ũ": 30
+}