Whisper: Fix decoder inputs for static pipeline #1469
+19
−9
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| @@ -25,6 +25,7 @@ class WhisperPipeline::StaticWhisperPipeline : public WhisperPipeline::WhisperPi | |||
|
|
|||
| private: | |||
| WhisperInitializedModels m_models; | |||
| std::shared_ptr<ov::Model> m_decoder_model; | |||
There was a problem hiding this comment.
why do we need to store model? once model is compiled, we need to release ov::Model to free memory consumed by its weights
There was a problem hiding this comment.
We can't compile this model until generate() called
There was a problem hiding this comment.
Well, I'd assume we have something like this:
class DecoderCache {
public:
ov::CompiledModel get_model(uint8_t input_id_size) {
// Get from hash table, otherwise compile and store...
}
private:
// [input_ids_size -> CompiledModel]
std::unordered_map<uint8_t, ov::CompiledModel> m_cache;
std::shared_ptr<ov::Model> decoder_model; // <- this is dynamic w/o transformation applied
}
// Whenever we need a model:
auto decoder = m_decoder_cache.get(input_ids_size);
| std::fill(input_ids_data + init_ids.size(), | ||
| input_ids_data + input_ids_tensor.get_size(), | ||
| static_cast<int32_t>(pad_token)); | ||
| // std::fill(input_ids_data + init_ids.size(), |
| std::fill(attention_mask_ptr, attention_mask_ptr + 3u, 0); | ||
| std::fill(attention_mask_ptr + 3u, attention_mask_ptr + attention_mask.get_size() - 2, 1); | ||
| std::fill(attention_mask_ptr, attention_mask_ptr + init_ids_size, 0); | ||
| std::fill(attention_mask_ptr + init_ids_size, attention_mask_ptr + attention_mask.get_size() - 2, 1); |
There was a problem hiding this comment.
Btw, since we reshape model anyway, we don't need attention_mask at all, probably we may not apply transformation that expose this
| @@ -489,7 +489,7 @@ void preprocess_decoder(std::shared_ptr<ov::Model> model) { | |||
| preprocessor.input("attention_mask").preprocess().convert_element_type(); | |||
| } else if (tensor.get_any_name().find("encoder_hidden_states") != std::string::npos) { | |||
| preprocessor.input("encoder_hidden_states").tensor().set_element_type(ov::element::Type_t::f16); | |||
| preprocessor.input("encoder_hidden_states").preprocess().convert_element_type(ov::element::Type_t::f32); // () | |||
| @@ -654,7 +657,13 @@ WhisperDecodedResults WhisperPipeline::StaticWhisperPipeline::generate( | |||
|
|
|||
| // prepare init_ids just once for whole input | |||
| if (init_ids.empty()) { | |||
| OPENVINO_ASSERT(m_models.decoder.get_tensor("input_ids").get_shape().back() == 1); | |||
There was a problem hiding this comment.
What does this check do?
dmatveev
changed the title
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Tickets: