Upload dataset

README.md

@@ -5,6 +5,10 @@ configs:
   data_files:
   - split: eval
     path: data/eval-*
+- config_name: scores
+  data_files:
+  - split: eval
+    path: scores/eval-*
 - config_name: transcripts
   data_files:
   - split: eval
@@ -27,6 +31,100 @@ dataset_info:
   splits:
   - name: eval
     num_examples: 10
+- config_name: scores
+  features:
+  - name: id
+    dtype: string
+  - name: variant
+    dtype: string
+  - name: ref_len
+    dtype: float64
+  - name: WER_parakeet
+    dtype: float64
+  - name: WER_parakeet_insertions
+    dtype: float64
+  - name: WER_parakeet_deletions
+    dtype: float64
+  - name: WER_parakeet_substitutions
+    dtype: float64
+  - name: WER_parakeet_n_ref_words
+    dtype: float64
+  - name: WER_parakeet_cpp_lib
+    dtype: float64
+  - name: WER_parakeet_cpp_lib_insertions
+    dtype: float64
+  - name: WER_parakeet_cpp_lib_deletions
+    dtype: float64
+  - name: WER_parakeet_cpp_lib_substitutions
+    dtype: float64
+  - name: WER_parakeet_cpp_lib_n_ref_words
+    dtype: float64
+  - name: WER_parakeet_cpp_lib_tdt_0_6b_v3_f16_20260814
+    dtype: float64
+  - name: WER_parakeet_cpp_lib_tdt_0_6b_v3_f16_20260814_insertions
+    dtype: float64
+  - name: WER_parakeet_cpp_lib_tdt_0_6b_v3_f16_20260814_deletions
+    dtype: float64
+  - name: WER_parakeet_cpp_lib_tdt_0_6b_v3_f16_20260814_substitutions
+    dtype: float64
+  - name: WER_parakeet_cpp_lib_tdt_0_6b_v3_f16_20260814_n_ref_words
+    dtype: float64
+  - name: WER_parakeet_nvidia_parakeet_tdt_0_6b_v3_20260814
+    dtype: float64
+  - name: WER_parakeet_nvidia_parakeet_tdt_0_6b_v3_20260814_insertions
+    dtype: float64
+  - name: WER_parakeet_nvidia_parakeet_tdt_0_6b_v3_20260814_deletions
+    dtype: float64
+  - name: WER_parakeet_nvidia_parakeet_tdt_0_6b_v3_20260814_substitutions
+    dtype: float64
+  - name: WER_parakeet_nvidia_parakeet_tdt_0_6b_v3_20260814_n_ref_words
+    dtype: float64
+  - name: WER_qwen
+    dtype: float64
+  - name: WER_qwen_insertions
+    dtype: float64
+  - name: WER_qwen_deletions
+    dtype: float64
+  - name: WER_qwen_substitutions
+    dtype: float64
+  - name: WER_qwen_n_ref_words
+    dtype: float64
+  - name: WER_qwen_Qwen_Qwen3_ASR_0_6B_20260814
+    dtype: float64
+  - name: WER_qwen_Qwen_Qwen3_ASR_0_6B_20260814_insertions
+    dtype: float64
+  - name: WER_qwen_Qwen_Qwen3_ASR_0_6B_20260814_deletions
+    dtype: float64
+  - name: WER_qwen_Qwen_Qwen3_ASR_0_6B_20260814_substitutions
+    dtype: float64
+  - name: WER_qwen_Qwen_Qwen3_ASR_0_6B_20260814_n_ref_words
+    dtype: float64
+  - name: WER_whisper_turbo
+    dtype: float64
+  - name: WER_whisper_turbo_insertions
+    dtype: float64
+  - name: WER_whisper_turbo_deletions
+    dtype: float64
+  - name: WER_whisper_turbo_substitutions
+    dtype: float64
+  - name: WER_whisper_turbo_n_ref_words
+    dtype: float64
+  - name: WER_whisper_turbo_turbo_20260814
+    dtype: float64
+  - name: WER_whisper_turbo_turbo_20260814_insertions
+    dtype: float64
+  - name: WER_whisper_turbo_turbo_20260814_deletions
+    dtype: float64
+  - name: WER_whisper_turbo_turbo_20260814_substitutions
+    dtype: float64
+  - name: WER_whisper_turbo_turbo_20260814_n_ref_words
+    dtype: float64
+  splits:
+  - name: eval
+    num_bytes: 3481
+    num_examples: 10
+  download_size: 27187
+  dataset_size: 3481
 - config_name: transcripts
   features:
   - name: id

scores/eval-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:19fffe6c3470af48681fc9dd06ed4111ed80626153cfeb0a9092ee6788519fd5
+size 27187