upload checkpoint

README.md

@@ -1,3 +1,21 @@
----
-license: cc-by-nc-sa-4.0
----
+---
+license: cc-by-nc-sa-4.0
+language:
+- de
+library_name: transformers
+pipeline_tag: automatic-speech-recognition
+base_model:
+- UsefulSensors/moonshine-tiny
+---
+# Moonshine
+
+This is a, to German, transfer learned moonshine tiny based on the English [UsefulSensors/moonshine-tiny](https://huggingface.co/UsefulSensors/moonshine-tiny)
+The new weights in this repository are licensed under [Attribution-NonCommercial-ShareAlike 4.0 International](https://creativecommons.org/licenses/by-nc-sa/4.0/).
+
+Original Inference Code: https://github.com/moonshine-ai/moonshine
+ONNX Rust Port: https://github.com/fidoriel/moonshine_onnx_rs
+
+## Eval
+Evaluation on Common Voice test split (https://huggingface.co/datasets/fidoriel/cv-22-de)
+WER: 0.114
+CER: 0.042

config.json

@@ -0,0 +1,33 @@
+{
+  "architectures": [
+    "MoonshineForConditionalGeneration"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "pad_token_id": 2,
+  "decoder_hidden_act": "silu",
+  "decoder_num_attention_heads": 8,
+  "decoder_num_hidden_layers": 6,
+  "decoder_num_key_value_heads": 8,
+  "decoder_start_token_id": 1,
+  "encoder_hidden_act": "gelu",
+  "encoder_num_attention_heads": 8,
+  "encoder_num_hidden_layers": 6,
+  "encoder_num_key_value_heads": 8,
+  "eos_token_id": 2,
+  "hidden_size": 288,
+  "initializer_range": 0.02,
+  "intermediate_size": 1152,
+  "is_encoder_decoder": true,
+  "max_position_embeddings": 194,
+  "model_type": "moonshine",
+  "partial_rotary_factor": 0.9,
+  "rope_scaling": null,
+  "rope_theta": 10000.0,
+  "torch_dtype": "float32",
+  "transformers_version": "4.48.0.dev0",
+  "use_cache": true,
+  "vocab_size": 32768,
+  "pad_head_dim_to_multiple_of": 8
+}

generation_config.json

@@ -0,0 +1,9 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "decoder_start_token_id": 1,
+  "pad_token_id": 2,
+  "eos_token_id": 2,
+  "transformers_version": "4.48.0.dev0",
+  "max_length": 194
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c9eb2859f8fec545f3f17d04e4b9fe9ba2e4e1b60da6e3e7737172fb82b709bf
+size 108389192

preprocessor_config.json

@@ -0,0 +1,9 @@
+{
+  "do_normalize": false,
+  "feature_extractor_type": "Wav2Vec2FeatureExtractor",
+  "feature_size": 1,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "return_attention_mask": true,
+  "sampling_rate": 16000
+}