README.md

---
license: apache-2.0
language:
  - code
library_name: transformers
pipeline_tag: feature-extraction
base_model: codesage/codesage-large-v2
tags:
  - onnx
  - teradata
  - byom
  - embeddings
  - feature-extraction
---



> Read the disclaimer below before using this model.

----

# codesage-large-v2 -- ONNX for Teradata BYOM

This repository hosts an **ONNX-converted** version of the upstream
model [`codesage/codesage-large-v2`](https://huggingface.co/codesage/codesage-large-v2),
packaged for the Teradata Vantage `mldb.ONNXEmbeddings` BYOM
function. It is **not** the original PyTorch model -- only the
inference graph and tokenizer needed for in-database embedding
generation.

What's different from upstream:

- **Format**: ONNX (opset 14, IR version 8 -- BYOM 6+ compatible),
  produced from the upstream weights with architecture-aware
  post-processing baked in.
- **Precision**: dynamic int8 quantization. See the variants table
  below for what is shipped for this model.
- **Pooling and post-processing**: this graph emits the raw
  `sentence_embedding` tensor. Pooling rule is
  **mean**.
- **Verification**: every variant's cosine fidelity vs. the
  upstream PyTorch reference is recorded on a fixed
  CodeSearchNet sample. Numbers may not generalize
  to your data.

## Model details

| | |
|---|---|
| Upstream repo | [`codesage/codesage-large-v2`](https://huggingface.co/codesage/codesage-large-v2) |
| Architecture | `CodeSage` (encoder) |
| Parameters | 1,313,464,320 |
| Output dimensions | 2048 |
| Pooling | `mean` |
| Instruction prefix | no |
| Max input tokens (advertised) | 2048 |
| Languages | 9 |
| License | apache-2.0 |
| ONNX opset | 14 |
| ONNX IR version | 8 (BYOM 6+ compatible) |

<details>
<summary>Full language list (9)</summary>

- `c`
- `c-sharp`
- `go`
- `java`
- `javascript`
- `typescript`
- `php`
- `python`
- `ruby`

</details>

## Quantization variants

This repository ships the following variants. Quality numbers are
measured against the upstream PyTorch reference on a fixed
CodeSearchNet sample. The **Size** column is the
on-disk size of the ONNX weight file in megabytes (MB, 10^6 bytes).

| Variant | Size (MB) | p50 cosine | R@1 |
|---|---|---|---|
| `ffn_skip` | 1318.9 | 0.819499 | 0.919 |


How to read the quality columns:

- **p50 cosine** is the median cosine similarity between this
  variant's embeddings and the fp32 ONNX reference, computed over
  a fixed evaluation set. Higher means closer to the unquantized
  model; **1.0** is identical.
- **R@1** is top-1 retrieval consistency: if you use this variant
  as a search index, R@1 is the fraction of queries that get the
  same nearest neighbor as the fp32 reference would. Higher is
  better.

Notes:
- **ffn_skip**: dynamic int8 with the feed-forward (FFN) MatMul
  layers kept in **fp32**, while attention and projection MatMuls
  stay quantized. The FFN layers are where most of the quantization
  error in transformer blocks concentrates; leaving them in fp32
  recovers most of the quality loss for a modest size increase.
  The artifact is roughly **3x smaller than fp32** (larger than the
  per_channel int8 sibling).

## Quickstart: using this model with Teradata BYOM

Requires Teradata Vantage with **BYOM 6+** (`mldb.ONNXEmbeddings`).

```python
import getpass
import teradataml as tdml
from huggingface_hub import hf_hub_download

repo_id   = "Teradata/codesage-large-v2"
model_id  = "codesage-large-v2"        # arbitrary, used as the BYOM model_id
onnx_file = "onnx/model-ffn_skip.onnx"

# 1. Download the ONNX + tokenizer for the chosen variant.
hf_hub_download(repo_id=repo_id, filename=onnx_file,       local_dir="./")
hf_hub_download(repo_id=repo_id, filename="tokenizer.json", local_dir="./")

# 2. Connect to Vantage.
tdml.create_context(
    host=input("host: "),
    username=input("user: "),
    password=getpass.getpass("password: "),
)

# 3. Load model + tokenizer into BYOM tables (one-time per model_id).
tdml.save_byom(model_id=model_id, model_file=onnx_file,
               table_name="embeddings_models")
tdml.save_byom(model_id=model_id, model_file="tokenizer.json",
               table_name="embeddings_tokenizers")
```

Then call `mldb.ONNXEmbeddings` against an input table whose
`txt` column carries the strings to embed:

This model emits a **2048-dimensional** embedding.
Teradata's wide-table output projection is capped at **2048 columns**,
which blocks the `FLOAT32(2048) + Accumulate('id')`
projection that smaller models use. Pick the SQL form that matches
your Vantage version:

**Option A -- `VARBYTE` (works on TD 17.20 and TD 20.0+)**

The vector lands as raw header-less float32 bytes
(`2048 * 4 = 8192` bytes),
which fits in a single `VARBYTE` column and dodges the 2048-column cap.
Decode on the client side or wrap the call in a UDF that returns
`FLOAT[]`.

```sql
SELECT *
FROM mldb.ONNXEmbeddings(
    ON (SELECT id, txt FROM your_input_table) AS InputTable
    ON (SELECT model_id, model FROM embeddings_models
         WHERE model_id = 'codesage-large-v2') AS ModelTable DIMENSION
    ON (SELECT model_id, tokenizer FROM embeddings_tokenizers
         WHERE model_id = 'codesage-large-v2') AS TokenizerTable DIMENSION
    USING
        Accumulate('id')
        ModelOutputTensor('sentence_embedding')
        OutputFormat('VARBYTE(8192)')
        OverwriteCachedModel('*')
) AS t
ORDER BY id;
```

**Option B -- `VECTOR` (TD 20.0+ only)**

Vantage 20.0 introduced a native `VECTOR` datatype that holds the
full embedding as a single typed column, with native vector-similarity
operators available on it.

```sql
SELECT *
FROM mldb.ONNXEmbeddings(
    ON (SELECT id, txt FROM your_input_table) AS InputTable
    ON (SELECT model_id, model FROM embeddings_models
         WHERE model_id = 'codesage-large-v2') AS ModelTable DIMENSION
    ON (SELECT model_id, tokenizer FROM embeddings_tokenizers
         WHERE model_id = 'codesage-large-v2') AS TokenizerTable DIMENSION
    USING
        Accumulate('id')
        ModelOutputTensor('sentence_embedding')
        OutputFormat('VECTOR')
        OverwriteCachedModel('*')
) AS t
ORDER BY id;
```

Use `VECTOR` if your Vantage version supports it; otherwise fall back
to `VARBYTE`. Both forms emit the same underlying float32 values.

Pooling rule **`mean`** is applied **inside** the converted
ONNX graph -- the output tensor named above already contains the
pooled, post-processed embedding vector.

## Original model attribution

The original weights and training methodology belong to
**the CodeSage authors**. Please cite their work, not this
repository, in academic contexts. The canonical upstream model card
is at
[`codesage/codesage-large-v2`](https://huggingface.co/codesage/codesage-large-v2);
refer to it for benchmarks, training details, intended use, and
citation information.

## Reporting issues

For ONNX-conversion or BYOM-compatibility issues specific to this
Teradata-converted artifact, please open a **Discussion** on this
model's Hugging Face page. Questions about the underlying model
quality, training, or intended use should go to the upstream
maintainer's model card.

----

DISCLAIMER: The content herein ("Content") is provided "AS IS" and is not covered by any Teradata Operations, Inc. and its affiliates ("Teradata") agreements. Its listing here does not constitute certification or endorsement by Teradata.

To the extent any of the Content contains or is related to any artificial intelligence ("AI") or other language learning models ("Models") that interoperate with the products and services of Teradata, by accessing, bringing, deploying or using such Models, you acknowledge and agree that you are solely responsible for ensuring compliance with all applicable laws, regulations, and restrictions governing the use, deployment, and distribution of AI technologies. This includes, but is not limited to, AI Diffusion Rules, European Union AI Act, AI-related laws and regulations, privacy laws, export controls, and financial or sector-specific regulations.

While Teradata may provide support, guidance, or assistance in the deployment or implementation of Models to interoperate with Teradata's products and/or services, you remain fully responsible for ensuring that your Models, data, and applications comply with all relevant legal and regulatory obligations. Our assistance does not constitute legal or regulatory approval, and Teradata disclaims any liability arising from non-compliance with applicable laws.

You must determine the suitability of the Models for any purpose. Given the probabilistic nature of machine learning and modeling, the use of the Models may in some situations result in incorrect output that does not accurately reflect the action generated. You should evaluate the accuracy of any output as appropriate for your use case, including by using human review of the output.