Upload folder using huggingface_hub

.ipynb_checkpoints/README-checkpoint.md

@@ -0,0 +1,100 @@
+---
+language:
+- en
+- tr
+tags:
+- translation
+license: cc-by-4.0
+datasets:
+- quickmt/quickmt-train.tr-en
+model-index:
+- name: quickmt-tr-en
+  results:
+  - task:
+      name: Translation tur-eng
+      type: translation
+      args: tur-eng
+    dataset:
+      name: flores101-devtest
+      type: flores_101
+      args: tur_Latn eng_Latn devtest
+    metrics:
+    - name: BLEU
+      type: bleu
+      value: 39.5
+    - name: CHRF
+      type: chrf
+      value: 65.03
+    - name: COMET
+      type: comet
+      value: 89.0
+---
+
+
+# `quickmt-tr-en` Neural Machine Translation Model 
+
+`quickmt-tr-en` is a reasonably fast and reasonably accurate neural machine translation model for translation from `tr` into `en`.
+
+
+## Model Information
+
+* Trained using [`eole`](https://github.com/eole-nlp/eole)
+* 185M parameter transformer 'big' with 8 encoder layers and 2 decoder layers
+* 50k joint Sentencepiece vocabulary
+* Expested for fast inference to [CTranslate2](https://github.com/OpenNMT/CTranslate2) format
+* Training data: https://huggingface.co/datasets/quickmt/quickmt-train.tr-en/tree/main
+
+See the `eole` model configuration in this repository for further details and the `eole-model` for the raw `eole` (pytorch) model.
+
+## Usage with `quickmt`
+
+You must install the Nvidia cuda toolkit first, if you want to do GPU inference.
+
+Next, install the `quickmt` python library and download the model:
+
+```bash
+git clone https://github.com/quickmt/quickmt.git
+pip install ./quickmt/
+
+quickmt-model-download quickmt/quickmt-tr-en ./quickmt-tr-en
+```
+
+Finally use the model in python:
+
+```python
+from quickmt impest Translator
+
+# Auto-detects GPU, set to "cpu" to force CPU inference
+t = Translator("./quickmt-tr-en/", device="auto")
+
+# Translate - set beam size to 1 for faster speed (but lower quality)
+sample_text = "Halifax, Nova Scotia'daki Dalhousie Üniversitesi'nde tıp profesörü ve Kanada Diyabet Derneği'nin klinik ve bilim alanı başkanı olan Dr. Ehud Ur, çalışmanın henüz başlangıç aşamasında olduğu konusunda uyarıda bulundu."
+
+t(sample_text, beam_size=5)
+```
+
+> 'Dr. Ehud Ur, a professor of medicine at Dalhousie University in Halifax, Nova Scotia, and head of clinical and science at the Canadian Diabetes Association, warned that the study is still in its infancy.'
+
+```python
+# Get alternative translations by sampling
+# You can pass any cTranslate2 `translate_batch` arguments
+t([sample_text], sampling_temperature=1.2, beam_size=1, sampling_topk=50, sampling_topp=0.9)
+```
+
+> 'Dr. Ehud Ur, a professor of medicine at Dalhousie University in Halifax, Nova Scotia, and chief clinical and science officer of the Canadian Diabetes Society, warned that the study is still at its preliminary stage.'
+
+The model is in `ctranslate2` format, and the tokenizers are `sentencepiece`, so you can use `ctranslate2` directly instead of through `quickmt`. It is also possible  to get this model to work with e.g. [LibreTranslate](https://libretranslate.com/) which also uses `ctranslate2` and `sentencepiece`.
+
+## Metrics
+
+`bleu` and `chrf2` are calculated with [sacrebleu](https://github.com/mjpost/sacrebleu) on the [Flores200 `devtest` test set](https://huggingface.co/datasets/facebook/flores) ("tur_Latn"->"eng_Latn"). `comet22` with the [`comet`](https://github.com/Unbabel/COMET) library and the [default model](https://huggingface.co/Unbabel/wmt22-comet-da). "Time (s)" is the time in seconds to translate the flores-devtest dataset (1012 sentences) on an RTX 4070s GPU with batch size 32 (faster speed is possible using a larger batch size).
+
+
+|                                  |   bleu |   chrf2 |   comet22 |   Time (s) |
+|:---------------------------------|-------:|--------:|----------:|-----------:|
+| quickmt/quickmt-tr-en            |  39.5  |   65.03 |     89    |       1.28 |
+| Helsinki-NLP/opus-mt-tr-en       |  32.19 |   59.43 |     87.03 |       3.5  |
+| facebook/nllb-200-distilled-600M |  37.15 |   62.76 |     88.28 |      20.91 |
+| facebook/nllb-200-distilled-1.3B |  39.33 |   64.88 |     89.27 |      36.35 |
+| facebook/m2m100_418M             |  28.53 |   57.08 |     84.59 |      17.34 |
+| facebook/m2m100_1.2B             |  33.1  |   60.38 |     86.82 |      34.1  |

.ipynb_checkpoints/eole-config-checkpoint.yaml

@@ -0,0 +1,96 @@
+## IO
+save_data: data
+overwrite: True
+seed: 1234
+report_every: 100
+valid_metrics: ["BLEU"]
+tensorboard: true
+tensorboard_log_dir: tensorboard
+
+### Vocab
+src_vocab: tr.eole.vocab
+tgt_vocab: en.eole.vocab
+src_vocab_size: 20000
+tgt_vocab_size: 20000
+vocab_size_multiple: 8
+share_vocab: false
+n_sample: 0
+
+data:
+    corpus_1:
+        path_src: hf://quickmt/quickmt-train.tr-en/tr
+        path_tgt: hf://quickmt/quickmt-train.tr-en/en
+        path_sco: hf://quickmt/quickmt-train.tr-en/sco
+    valid:
+        path_src: valid.tr
+        path_tgt: valid.en
+
+transforms: [sentencepiece, filtertoolong]
+transforms_configs:
+  sentencepiece:
+    src_subword_model: "tr.spm.model"
+    tgt_subword_model: "en.spm.model"
+  filtertoolong:
+    src_seq_length: 256
+    tgt_seq_length: 256
+
+training:
+    # Run configuration
+    model_path: quickmt-tr-en-eole-model
+    #train_from: model
+    keep_checkpoint: 4
+    train_steps: 100000
+    save_checkpoint_steps: 5000
+    valid_steps: 5000
+    
+    # Train on a single GPU
+    world_size: 1
+    gpu_ranks: [0]
+
+    # Batching 10240
+    batch_type: "tokens"
+    batch_size: 8000
+    valid_batch_size: 4096
+    batch_size_multiple: 8
+    accum_count: [10]
+    accum_steps: [0]
+
+    # Optimizer & Compute
+    compute_dtype: "fp16"
+    optim: "adamw"
+    #use_amp: False
+    learning_rate: 2.0
+    warmup_steps: 4000
+    decay_method: "noam"
+    adam_beta2: 0.998
+
+    # Data loading
+    bucket_size: 128000
+    num_workers: 4
+    prefetch_factor: 32
+    
+    # Hyperparams
+    dropout_steps: [0]
+    dropout: [0.1]
+    attention_dropout: [0.1]
+    max_grad_norm: 0
+    label_smoothing: 0.1
+    average_decay: 0.0001
+    param_init_method: xavier_uniform
+    normalization: "tokens"
+
+model:
+    architecture: "transformer"
+    share_embeddings: false
+    share_decoder_embeddings: false
+    hidden_size: 1024
+    encoder:
+        layers: 8
+    decoder:
+        layers: 2
+    heads: 8
+    transformer_ff: 4096
+    embeddings:
+        word_vec_size: 1024
+        position_encoding_type: "SinusoidalInterleaved"
+

README.md

@@ -1,3 +1,100 @@
----
-license: cc-by-4.0
----
+---
+language:
+- en
+- tr
+tags:
+- translation
+license: cc-by-4.0
+datasets:
+- quickmt/quickmt-train.tr-en
+model-index:
+- name: quickmt-tr-en
+  results:
+  - task:
+      name: Translation tur-eng
+      type: translation
+      args: tur-eng
+    dataset:
+      name: flores101-devtest
+      type: flores_101
+      args: tur_Latn eng_Latn devtest
+    metrics:
+    - name: BLEU
+      type: bleu
+      value: 39.5
+    - name: CHRF
+      type: chrf
+      value: 65.03
+    - name: COMET
+      type: comet
+      value: 89.0
+---
+
+
+# `quickmt-tr-en` Neural Machine Translation Model 
+
+`quickmt-tr-en` is a reasonably fast and reasonably accurate neural machine translation model for translation from `tr` into `en`.
+
+
+## Model Information
+
+* Trained using [`eole`](https://github.com/eole-nlp/eole)
+* 185M parameter transformer 'big' with 8 encoder layers and 2 decoder layers
+* 50k joint Sentencepiece vocabulary
+* Expested for fast inference to [CTranslate2](https://github.com/OpenNMT/CTranslate2) format
+* Training data: https://huggingface.co/datasets/quickmt/quickmt-train.tr-en/tree/main
+
+See the `eole` model configuration in this repository for further details and the `eole-model` for the raw `eole` (pytorch) model.
+
+## Usage with `quickmt`
+
+You must install the Nvidia cuda toolkit first, if you want to do GPU inference.
+
+Next, install the `quickmt` python library and download the model:
+
+```bash
+git clone https://github.com/quickmt/quickmt.git
+pip install ./quickmt/
+
+quickmt-model-download quickmt/quickmt-tr-en ./quickmt-tr-en
+```
+
+Finally use the model in python:
+
+```python
+from quickmt impest Translator
+
+# Auto-detects GPU, set to "cpu" to force CPU inference
+t = Translator("./quickmt-tr-en/", device="auto")
+
+# Translate - set beam size to 1 for faster speed (but lower quality)
+sample_text = "Halifax, Nova Scotia'daki Dalhousie Üniversitesi'nde tıp profesörü ve Kanada Diyabet Derneği'nin klinik ve bilim alanı başkanı olan Dr. Ehud Ur, çalışmanın henüz başlangıç aşamasında olduğu konusunda uyarıda bulundu."
+
+t(sample_text, beam_size=5)
+```
+
+> 'Dr. Ehud Ur, a professor of medicine at Dalhousie University in Halifax, Nova Scotia, and head of clinical and science at the Canadian Diabetes Association, warned that the study is still in its infancy.'
+
+```python
+# Get alternative translations by sampling
+# You can pass any cTranslate2 `translate_batch` arguments
+t([sample_text], sampling_temperature=1.2, beam_size=1, sampling_topk=50, sampling_topp=0.9)
+```
+
+> 'Dr. Ehud Ur, a professor of medicine at Dalhousie University in Halifax, Nova Scotia, and chief clinical and science officer of the Canadian Diabetes Society, warned that the study is still at its preliminary stage.'
+
+The model is in `ctranslate2` format, and the tokenizers are `sentencepiece`, so you can use `ctranslate2` directly instead of through `quickmt`. It is also possible  to get this model to work with e.g. [LibreTranslate](https://libretranslate.com/) which also uses `ctranslate2` and `sentencepiece`.
+
+## Metrics
+
+`bleu` and `chrf2` are calculated with [sacrebleu](https://github.com/mjpost/sacrebleu) on the [Flores200 `devtest` test set](https://huggingface.co/datasets/facebook/flores) ("tur_Latn"->"eng_Latn"). `comet22` with the [`comet`](https://github.com/Unbabel/COMET) library and the [default model](https://huggingface.co/Unbabel/wmt22-comet-da). "Time (s)" is the time in seconds to translate the flores-devtest dataset (1012 sentences) on an RTX 4070s GPU with batch size 32 (faster speed is possible using a larger batch size).
+
+
+|                                  |   bleu |   chrf2 |   comet22 |   Time (s) |
+|:---------------------------------|-------:|--------:|----------:|-----------:|
+| quickmt/quickmt-tr-en            |  39.5  |   65.03 |     89    |       1.28 |
+| Helsinki-NLP/opus-mt-tr-en       |  32.19 |   59.43 |     87.03 |       3.5  |
+| facebook/nllb-200-distilled-600M |  37.15 |   62.76 |     88.28 |      20.91 |
+| facebook/nllb-200-distilled-1.3B |  39.33 |   64.88 |     89.27 |      36.35 |
+| facebook/m2m100_418M             |  28.53 |   57.08 |     84.59 |      17.34 |
+| facebook/m2m100_1.2B             |  33.1  |   60.38 |     86.82 |      34.1  |

config.json

@@ -0,0 +1,10 @@
+{
+  "add_source_bos": false,
+  "add_source_eos": false,
+  "bos_token": "<s>",
+  "decoder_start_token": "<s>",
+  "eos_token": "</s>",
+  "layer_norm_epsilon": 1e-06,
+  "multi_query_attention": false,
+  "unk_token": "<unk>"
+}

eole-config.yaml

@@ -0,0 +1,96 @@
+## IO
+save_data: data
+overwrite: True
+seed: 1234
+report_every: 100
+valid_metrics: ["BLEU"]
+tensorboard: true
+tensorboard_log_dir: tensorboard
+
+### Vocab
+src_vocab: tr.eole.vocab
+tgt_vocab: en.eole.vocab
+src_vocab_size: 20000
+tgt_vocab_size: 20000
+vocab_size_multiple: 8
+share_vocab: false
+n_sample: 0
+
+data:
+    corpus_1:
+        path_src: hf://quickmt/quickmt-train.tr-en/tr
+        path_tgt: hf://quickmt/quickmt-train.tr-en/en
+        path_sco: hf://quickmt/quickmt-train.tr-en/sco
+    valid:
+        path_src: valid.tr
+        path_tgt: valid.en
+
+transforms: [sentencepiece, filtertoolong]
+transforms_configs:
+  sentencepiece:
+    src_subword_model: "tr.spm.model"
+    tgt_subword_model: "en.spm.model"
+  filtertoolong:
+    src_seq_length: 256
+    tgt_seq_length: 256
+
+training:
+    # Run configuration
+    model_path: quickmt-tr-en-eole-model
+    #train_from: model
+    keep_checkpoint: 4
+    train_steps: 100000
+    save_checkpoint_steps: 5000
+    valid_steps: 5000
+    
+    # Train on a single GPU
+    world_size: 1
+    gpu_ranks: [0]
+
+    # Batching 10240
+    batch_type: "tokens"
+    batch_size: 8000
+    valid_batch_size: 4096
+    batch_size_multiple: 8
+    accum_count: [10]
+    accum_steps: [0]
+
+    # Optimizer & Compute
+    compute_dtype: "fp16"
+    optim: "adamw"
+    #use_amp: False
+    learning_rate: 2.0
+    warmup_steps: 4000
+    decay_method: "noam"
+    adam_beta2: 0.998
+
+    # Data loading
+    bucket_size: 128000
+    num_workers: 4
+    prefetch_factor: 32
+    
+    # Hyperparams
+    dropout_steps: [0]
+    dropout: [0.1]
+    attention_dropout: [0.1]
+    max_grad_norm: 0
+    label_smoothing: 0.1
+    average_decay: 0.0001
+    param_init_method: xavier_uniform
+    normalization: "tokens"
+
+model:
+    architecture: "transformer"
+    share_embeddings: false
+    share_decoder_embeddings: false
+    hidden_size: 1024
+    encoder:
+        layers: 8
+    decoder:
+        layers: 2
+    heads: 8
+    transformer_ff: 4096
+    embeddings:
+        word_vec_size: 1024
+        position_encoding_type: "SinusoidalInterleaved"
+

eole-model/config.json

@@ -0,0 +1,132 @@
+{
+  "n_sample": 0,
+  "valid_metrics": [
+    "BLEU"
+  ],
+  "tensorboard": true,
+  "tgt_vocab_size": 20000,
+  "save_data": "data",
+  "src_vocab_size": 20000,
+  "overwrite": true,
+  "seed": 1234,
+  "src_vocab": "tr.eole.vocab",
+  "tgt_vocab": "en.eole.vocab",
+  "transforms": [
+    "sentencepiece",
+    "filtertoolong"
+  ],
+  "vocab_size_multiple": 8,
+  "tensorboard_log_dir_dated": "tensorboard/Jul-21_11-21-32",
+  "tensorboard_log_dir": "tensorboard",
+  "share_vocab": false,
+  "report_every": 100,
+  "training": {
+    "decay_method": "noam",
+    "world_size": 1,
+    "optim": "adamw",
+    "param_init_method": "xavier_uniform",
+    "save_checkpoint_steps": 5000,
+    "batch_size": 8000,
+    "warmup_steps": 4000,
+    "batch_type": "tokens",
+    "normalization": "tokens",
+    "attention_dropout": [
+      0.1
+    ],
+    "compute_dtype": "torch.float16",
+    "gpu_ranks": [
+      0
+    ],
+    "max_grad_norm": 0.0,
+    "bucket_size": 128000,
+    "valid_batch_size": 4096,
+    "dropout_steps": [
+      0
+    ],
+    "learning_rate": 2.0,
+    "label_smoothing": 0.1,
+    "num_workers": 0,
+    "valid_steps": 5000,
+    "average_decay": 0.0001,
+    "accum_count": [
+      10
+    ],
+    "adam_beta2": 0.998,
+    "prefetch_factor": 32,
+    "model_path": "quickmt-tr-en-eole-model",
+    "accum_steps": [
+      0
+    ],
+    "train_steps": 100000,
+    "batch_size_multiple": 8,
+    "keep_checkpoint": 4,
+    "dropout": [
+      0.1
+    ]
+  },
+  "transforms_configs": {
+    "filtertoolong": {
+      "tgt_seq_length": 256,
+      "src_seq_length": 256
+    },
+    "sentencepiece": {
+      "tgt_subword_model": "${MODEL_PATH}/en.spm.model",
+      "src_subword_model": "${MODEL_PATH}/tr.spm.model"
+    }
+  },
+  "data": {
+    "corpus_1": {
+      "path_src": "train.tr",
+      "transforms": [
+        "sentencepiece",
+        "filtertoolong"
+      ],
+      "path_align": null,
+      "path_tgt": "train.en"
+    },
+    "valid": {
+      "path_src": "valid.tr",
+      "transforms": [
+        "sentencepiece",
+        "filtertoolong"
+      ],
+      "path_align": null,
+      "path_tgt": "valid.en"
+    }
+  },
+  "model": {
+    "heads": 8,
+    "hidden_size": 1024,
+    "share_decoder_embeddings": false,
+    "position_encoding_type": "SinusoidalInterleaved",
+    "transformer_ff": 4096,
+    "architecture": "transformer",
+    "share_embeddings": false,
+    "encoder": {
+      "encoder_type": "transformer",
+      "heads": 8,
+      "n_positions": null,
+      "src_word_vec_size": 1024,
+      "hidden_size": 1024,
+      "position_encoding_type": "SinusoidalInterleaved",
+      "layers": 8,
+      "transformer_ff": 4096
+    },
+    "embeddings": {
+      "word_vec_size": 1024,
+      "position_encoding_type": "SinusoidalInterleaved",
+      "tgt_word_vec_size": 1024,
+      "src_word_vec_size": 1024
+    },
+    "decoder": {
+      "n_positions": null,
+      "heads": 8,
+      "decoder_type": "transformer",
+      "hidden_size": 1024,
+      "position_encoding_type": "SinusoidalInterleaved",
+      "tgt_word_vec_size": 1024,
+      "layers": 2,
+      "transformer_ff": 4096
+    }
+  }
+}

eole-model/en.spm.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c4b6f8bf657c61b0219015022eba70ae06f19c7ae4ac9bd7dbacdb9bafa2ea4
+size 586127

eole-model/model.00.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8820d106623bf3225e7527a3a2e78ebdd9bec39c8e1a54de668233d68a8dff32
+size 823882912

eole-model/tr.spm.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:453f064230753d99aecab739b8de017e91836692acd2566812de6f96ffa83f26
+size 604393

model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7151fa84b90fe90dc99dae1f7bba3c0b416d3d70cc3ce0ae80f60c543c625745
+size 401699775

src.spm.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:453f064230753d99aecab739b8de017e91836692acd2566812de6f96ffa83f26
+size 604393

tgt.spm.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c4b6f8bf657c61b0219015022eba70ae06f19c7ae4ac9bd7dbacdb9bafa2ea4
+size 586127