update doc

Author: hydropix

README.md

@@ -104,8 +104,11 @@ python translate.py -i book.txt --provider openai \
 | `-tl, --target_lang` | Target language | Chinese |
 | `-m, --model` | Model name | mistral-small:24b |
 | `--provider` | ollama/openrouter/openai/gemini | ollama |
+| `--text-cleanup` | OCR/typographic cleanup | disabled |
+| `--refine` | Second pass for literary polish | disabled |
+| `--tts` | Generate audio (Edge-TTS) | disabled |
 
-See [docs/CLI.md](docs/CLI.md) for all options and examples.
+See [docs/CLI.md](docs/CLI.md) for all options (TTS voices, rates, formats, etc.).
 
 ---
 

docs/CLI.md

@@ -49,13 +49,22 @@ python translate.py -i input_file -o output_file
 | `--openai_api_key` | OpenAI API key |
 | `--gemini_api_key` | Gemini API key |
 
-### Performance
+### Prompt Options
+
+| Option | Description |
+|--------|-------------|
+| `--text-cleanup` | Enable OCR/typographic cleanup (fix broken lines, spacing, punctuation) |
+| `--refine` | Enable refinement pass: runs a second pass to polish translation quality and literary style |
+
+### TTS (Text-to-Speech)
 
 | Option | Description | Default |
 |--------|-------------|---------|
-| `-cs, --chunksize` | Lines per chunk | 25 |
-| `--timeout` | Request timeout (seconds) | 900 |
-| `--context-window` | Context window size | 2048 |
+| `--tts` | Generate audio from translated text using Edge-TTS | disabled |
+| `--tts-voice` | TTS voice name | Auto-selected based on target language |
+| `--tts-rate` | Speech rate adjustment (e.g., `+10%`, `-20%`) | +0% |
+| `--tts-bitrate` | Audio bitrate (e.g., `64k`, `96k`) | 48k |
+| `--tts-format` | Audio output format: `opus` or `mp3` | opus |
 
 ### Display
 
@@ -114,17 +123,30 @@ python translate.py -i book.txt -o book_fr.txt \
     -m your-model
 ```
 
-### Performance Tuning
+### With Prompt Options
+
+```bash
+# OCR cleanup (fix broken lines, spacing from scanned documents)
+python translate.py -i scanned_book.txt -tl French --text-cleanup
+
+# Refinement pass for higher quality literary translation
+python translate.py -i novel.epub -tl French --refine
+
+# Both options combined
+python translate.py -i scanned_book.txt -tl French --text-cleanup --refine
+```
+
+### With TTS (Text-to-Speech)
 
 ```bash
-# Larger chunks for better context (needs more VRAM)
-python translate.py -i book.txt -o book_fr.txt -cs 50
+# Generate audio with auto-selected voice
+python translate.py -i book.txt -tl French --tts
 
-# Smaller chunks for limited hardware
-python translate.py -i book.txt -o book_fr.txt -cs 15
+# Specify voice and format
+python translate.py -i book.txt -tl French --tts --tts-voice fr-FR-DeniseNeural --tts-format mp3
 
-# Longer timeout for slow models
-python translate.py -i book.txt -o book_fr.txt --timeout 1800
+# Adjust speech rate and quality
+python translate.py -i book.txt -tl French --tts --tts-rate "+10%" --tts-bitrate 96k
 ```
 
 ---
@@ -151,6 +173,13 @@ MAX_TOKENS_PER_CHUNK=400  # Token-based chunking (default: 400 tokens)
 # Languages
 DEFAULT_SOURCE_LANGUAGE=English
 DEFAULT_TARGET_LANGUAGE=French
+
+# TTS
+TTS_ENABLED=false
+TTS_VOICE=               # Auto-selected if empty
+TTS_RATE=+0%
+TTS_BITRATE=48k
+TTS_OUTPUT_FORMAT=opus
 ```
 
 ---
@@ -161,11 +190,3 @@ DEFAULT_TARGET_LANGUAGE=French
 |------|---------|
 | 0 | Success |
 | 1 | Error (check console output) |
-
----
-
-## Output Location
-
-By default, translated files are saved in `translated_files/` directory.
-
-Configure with `OUTPUT_DIR` environment variable.