| ---
|
| library_name: transformers
|
| license: apache-2.0
|
| pipeline_tag: translation
|
| language:
|
| - en
|
| - es
|
| - fr
|
| - de
|
| - it
|
| - pt
|
| - hi
|
| - zh
|
| - ja
|
| - ru
|
| - ar
|
| - ml
|
| - ta
|
| - te
|
| - bn
|
| - ur
|
| - id
|
| - vi
|
| - ko
|
| - tr
|
| - nl
|
| - pl
|
| - sv
|
| - fi
|
| - da
|
| - no
|
| - cs
|
| - el
|
| - he
|
| - th
|
| - sw
|
| - uk
|
| - ro
|
| - hu
|
| - fa
|
| - my
|
| - km
|
| - ne
|
| - si
|
| - am
|
| - so
|
| - pa
|
| - gu
|
| - kn
|
| - mr
|
| - or
|
| - as
|
| - sa
|
| - sd
|
| - ps
|
| - aar
|
| - abk
|
| - ace
|
| - acm
|
| - acu
|
| - ady
|
| - afb
|
| - afh
|
| - afr
|
| - agr
|
| - ain
|
| - aka
|
| - ake
|
| - akl
|
| - aln
|
| - amh
|
| - amu
|
| - ang
|
| - aoz
|
| - apc
|
| - ara
|
| - arg
|
| - arq
|
| - ary
|
| - arz
|
| - ase
|
| - asm
|
| - ast
|
| - ava
|
| - ave
|
| - avk
|
| - awa
|
| - aym
|
| - aze
|
| - bak
|
| - bal
|
| - bam
|
| - ban
|
| - bar
|
| - bbc
|
| - bbj
|
| - bel
|
| - bem
|
| - ben
|
| - ber
|
| - bho
|
| - bih
|
| - bis
|
| - bjn
|
| - bod
|
| - bom
|
| - bos
|
| - bre
|
| - brx
|
| - bsn
|
| - bua
|
| - bug
|
| - bul
|
| - bvy
|
| - bzd
|
| - bzt
|
| - cak
|
| - cat
|
| - cbk
|
| - cdo
|
| - ceb
|
| - ces
|
| - cha
|
| - che
|
| - cho
|
| - chq
|
| - chr
|
| - chu
|
| - chv
|
| - cjk
|
| - cjp
|
| - ckb
|
| - cmn
|
| - cnh
|
| - cni
|
| - cop
|
| - cor
|
| - cos
|
| - cre
|
| - crh
|
| - crp
|
| - csb
|
| - cym
|
| - czn
|
| - dan
|
| - deu
|
| - dik
|
| - din
|
| - diq
|
| - div
|
| - dje
|
| - djk
|
| - dop
|
| - drt
|
| - dsb
|
| - dtp
|
| - dws
|
| - dyu
|
| - dzo
|
| - efi
|
| - egl
|
| - ell
|
| - eng
|
| - enm
|
| - epo
|
| - est
|
| - eus
|
| - ewe
|
| - ext
|
| - fao
|
| - fas
|
| - fij
|
| - fil
|
| - fin
|
| - fkv
|
| - fon
|
| - fra
|
| - frm
|
| - frp
|
| - frr
|
| - fry
|
| - ful
|
| - fur
|
| - fuv
|
| - gan
|
| - gbi
|
| - gbm
|
| - gcf
|
| - gil
|
| - gla
|
| - gle
|
| - glg
|
| - glv
|
| - gom
|
| - gos
|
| - got
|
| - grc
|
| - grn
|
| - gsw
|
| - guj
|
| - gvc
|
| - hai
|
| - hak
|
| - hat
|
| - hau
|
| - haw
|
| - hbo
|
| - hbs
|
| - hch
|
| - heb
|
| - hif
|
| - hil
|
| - hin
|
| - hmo
|
| - hne
|
| - hoc
|
| - hrv
|
| - hrx
|
| - hsb
|
| - hun
|
| - hup
|
| - hus
|
| - hye
|
| - iba
|
| - ibo
|
| - ido
|
| - ike
|
| - iku
|
| - ile
|
| - ilo
|
| - ina
|
| - ind
|
| - inh
|
| - ipk
|
| - isl
|
| - ita
|
| - jak
|
| - jam
|
| - jav
|
| - jbo
|
| - jiv
|
| - jpn
|
| - kab
|
| - kac
|
| - kal
|
| - kam
|
| - kan
|
| - kas
|
| - kat
|
| - kau
|
| - kaz
|
| - kbh
|
| - kbp
|
| - kea
|
| - kek
|
| - kha
|
| - khm
|
| - kik
|
| - kin
|
| - kir
|
| - kjh
|
| - kmb
|
| - kmr
|
| - koi
|
| - kom
|
| - kon
|
| - kor
|
| - krl
|
| - ksh
|
| - kua
|
| - kur
|
| - lad
|
| - lao
|
| - lat
|
| - lav
|
| - ldn
|
| - lfn
|
| - lij
|
| - lim
|
| - lin
|
| - lit
|
| - liv
|
| - lkt
|
| - lld
|
| - lmo
|
| - ltg
|
| - ltz
|
| - lua
|
| - lug
|
| - luo
|
| - lus
|
| - lut
|
| - lzh
|
| - mad
|
| - mag
|
| - mah
|
| - mai
|
| - mal
|
| - mam
|
| - mar
|
| - max
|
| - mfe
|
| - mgm
|
| - mhr
|
| - mic
|
| - min
|
| - miq
|
| - mkd
|
| - mlg
|
| - mlt
|
| - mni
|
| - moh
|
| - mon
|
| - mos
|
| - mri
|
| - msa
|
| - mus
|
| - mvv
|
| - mwl
|
| - mww
|
| - mya
|
| - myv
|
| - nah
|
| - nan
|
| - nap
|
| - nau
|
| - nav
|
| - nbl
|
| - nch
|
| - nci
|
| - nds
|
| - nep
|
| - ngt
|
| - ngu
|
| - nhg
|
| - nij
|
| - niu
|
| - nld
|
| - nlv
|
| - nno
|
| - nob
|
| - nog
|
| - non
|
| - nor
|
| - nov
|
| - npi
|
| - nso
|
| - nst
|
| - nus
|
| - nya
|
| - oci
|
| - ojb
|
| - ood
|
| - ori
|
| - orm
|
| - orv
|
| - oss
|
| - ota
|
| - oto
|
| - pag
|
| - pam
|
| - pan
|
| - pap
|
| - pau
|
| - pck
|
| - pcm
|
| - pdc
|
| - pes
|
| - phn
|
| - pir
|
| - pli
|
| - plt
|
| - pms
|
| - pmy
|
| - pnb
|
| - pol
|
| - por
|
| - pot
|
| - ppk
|
| - ppl
|
| - prg
|
| - prs
|
| - pus
|
| - quc
|
| - que
|
| - quw
|
| - quy
|
| - quz
|
| - qya
|
| - rap
|
| - rar
|
| - rif
|
| - roh
|
| - rom
|
| - ron
|
| - rue
|
| - run
|
| - rup
|
| - rus
|
| - sag
|
| - sah
|
| - san
|
| - sat
|
| - scn
|
| - sco
|
| - sgs
|
| - shi
|
| - shn
|
| - shp
|
| - shs
|
| - shy
|
| - sin
|
| - sjn
|
| - slk
|
| - slv
|
| - sma
|
| - sme
|
| - sml
|
| - smo
|
| - sna
|
| - snd
|
| - som
|
| - sot
|
| - spa
|
| - sqi
|
| - srd
|
| - srp
|
| - ssw
|
| - stq
|
| - sun
|
| - sux
|
| - swa
|
| - swe
|
| - swg
|
| - swh
|
| - syr
|
| - szl
|
| - tah
|
| - tam
|
| - taq
|
| - tar
|
| - tat
|
| - tel
|
| - tet
|
| - tgk
|
| - tgl
|
| - tha
|
| - tir
|
| - tlh
|
| - tly
|
| - tmh
|
| - tmr
|
| - tok
|
| - ton
|
| - tpi
|
| - tpw
|
| - trv
|
| - tsn
|
| - tso
|
| - tuk
|
| - tum
|
| - tur
|
| - tvl
|
| - twi
|
| - tyj
|
| - tzl
|
| - tzm
|
| - udm
|
| - uig
|
| - ukr
|
| - umb
|
| - urd
|
| - usp
|
| - uzb
|
| - vec
|
| - ven
|
| - vie
|
| - vls
|
| - vol
|
| - wae
|
| - wal
|
| - war
|
| - wln
|
| - wol
|
| - wuu
|
| - xal
|
| - xho
|
| - xin
|
| - xmf
|
| - xzp
|
| - yaq
|
| - yid
|
| - yor
|
| - yue
|
| - zam
|
| - zen
|
| - zho
|
| - zhs
|
| - zht
|
| - zlm
|
| - zsm
|
| - zul
|
| - zza
|
| - ach
|
| - ada
|
| tags:
|
| - 500-languages
|
| - translation
|
| - multilingual
|
| - moe
|
| - mixture-of-experts
|
| - mythos2.0
|
| - seq2seq
|
| - 8k-context
|
| - document-packing
|
| - pytorch
|
| - open-weights
|
| - free-for-all
|
| widget:
|
| - text: "Artificial intelligence empowers communities around the world to connect and collaborate."
|
| example_title: "English to Spanish"
|
| output:
|
| text: "La inteligencia artificial permite a las comunidades de todo el mundo conectarse y colaborar."
|
| - text: "Artificial intelligence empowers communities around the world to connect and collaborate."
|
| example_title: "English to French"
|
| output:
|
| text: "L'intelligence artificielle permet aux communautés du monde entier de se connecter et de collaborer."
|
| - text: "Artificial intelligence empowers communities around the world to connect and collaborate."
|
| example_title: "English to German"
|
| output:
|
| text: "Künstliche Intelligenz ermöglicht es Gemeinschaften weltweit, sich zu verbinden und zusammenzuarbeiten."
|
| gated: auto
|
| extra_gated_prompt: "Access to Mythos 2.0 is granted automatically upon submission."
|
| extra_gated_fields:
|
| Username: text
|
| Email: text
|
| ---
|
|
|
| <div align="center">
|
|
|
| <img src="./figures/mythos_logo.svg" width="65%" style="max-width: 680px;" alt="Mythos Logo" />
|
|
|
| # Mythos2.0-2B
|
|
|
| **A Frontier Sparse Mixture-of-Experts (SMoE) Foundation Translation Model for 500+ Global Languages**
|
|
|
| <p>
|
| <a href="https://huggingface.co/collections/AdithyanAI"><img src="https://img.shields.io/badge/Collections-Mythos2.0_Series-007ACC?style=for-the-badge&logo=huggingface&logoColor=white" alt="Collections"></a>
|
| <a href="https://github.com/Adithyan435/Mythos-Multilingual-LLM"><img src="https://img.shields.io/badge/GitHub-Repository-181717?style=for-the-badge&logo=github&logoColor=white" alt="GitHub"></a>
|
| <a href="https://www.apache.org/licenses/LICENSE-2.0"><img src="https://img.shields.io/badge/License-Apache_2.0-D4AF37?style=for-the-badge&labelColor=0B0F19" alt="License"></a>
|
| </p>
|
|
|
|
|
| <p>
|
| <a href="#welcome-to-mythos-ai">Organization</a> ·
|
| <a href="#supported-languages-directory">Supported Languages (500+)</a> ·
|
| <a href="#mythos20-model-collections">Collections</a> ·
|
| <a href="#model-architecture-overview">Architecture</a> ·
|
| <a href="#quickstart-free-offline-inference">Quickstart</a> ·
|
| <a href="#community-and-collaboration">Collaboration</a>
|
| </p>
|
|
|
| </div>
|
|
|
| ---
|
|
|
| <a id="welcome-to-mythos-ai"></a>
|
| ## Welcome to Mythos AI
|
|
|
| **Mythos** is an open-source foundation model family built by **Adithyan AI**. In this organization, we develop and open-source state-of-the-art **Sparse Mixture-of-Experts (SMoE)** language models, universal translation engines, parallel multilingual datasets, and ultra-efficient inference runtimes targeting 500+ languages.
|
|
|
| * **Mission:** Bridge the digital divide for underserved languages worldwide through efficient, open-weights AI architectures.
|
| * **100% Free and Open-Source:** Released under the permissive **Apache 2.0** license with zero paywalls, metered tokens, or subscription fees.
|
| * **Community:** Explore our models and datasets on [Hugging Face](https://huggingface.co/AdithyanAI) and view development on [GitHub](https://github.com/Adithyan435/Mythos-Multilingual-LLM).
|
|
|
| ---
|
|
|
| <a id="supported-languages-directory"></a>
|
| ## 🌐 Supported Languages Directory (500+ Languages)
|
|
|
| Mythos2.0-2B was trained on a massive 16.5B sentence-pair parallel corpus covering **500+ languages and regional dialects across Africa, the Americas, Asia, Europe, and Oceania**.
|
|
|
| To translate any text into a desired language, simply prepend the target language tag `<2code>` (e.g. `<2es>` for Spanish, `<2ml>` for Malayalam, `<2hi>` for Hindi, `<2fr>` for French, `<2de>` for German, `<2ta>` for Tamil).
|
|
|
| ### Major Language Hubs Supported:
|
| * **Global Commercial Languages:** English (`eng`), Spanish (`spa`), French (`fra`), German (`deu`), Italian (`ita`), Portuguese (`por`), Russian (`rus`), Mandarin Chinese (`cmn`), Japanese (`jpn`), Korean (`kor`), Arabic (`ara`), Turkish (`tur`), Vietnamese (`vie`), Indonesian (`ind`), Dutch (`nld`), Polish (`pol`).
|
| * **South Asian & Indian Languages:** Hindi (`hin`), Malayalam (`mal`), Tamil (`tam`), Telugu (`tel`), Bengali (`ben`), Marathi (`mar`), Gujarati (`guj`), Kannada (`kan`), Punjabi (`pan`), Urdu (`urd`), Odia (`ori`), Assamese (`asm`), Sanskrit (`san`), Nepali (`nep`), Sinhala (`sin`), Maithili (`mai`), Bhojpuri (`bho`), Sindhi (`snd`), Kashmiri (`kas`), Konkani (`kok`).
|
| * **African Languages:** Swahili (`swa`), Amharic (`amh`), Yoruba (`yor`), Igbo (`ibo`), Hausa (`hau`), Somali (`som`), Oromo (`orm`), Zulu (`zul`), Xhosa (`xho`), Shona (`sna`), Tigrinya (`tir`), Malagasy (`mlg`), Kinyarwanda (`kin`), Lingala (`lin`), Bambara (`bam`), Wolof (`wol`).
|
| * **European & Slavic Languages:** Ukrainian (`ukr`), Czech (`ces`), Romanian (`ron`), Greek (`ell`), Hungarian (`hun`), Danish (`dan`), Finnish (`fin`), Norwegian (`nob`), Swedish (`swe`), Bulgarian (`bul`), Croatian (`hrv`), Serbian (`srp`), Slovak (`slk`), Catalan (`cat`), Basque (`eus`), Galician (`glg`), Irish (`gle`), Welsh (`cym`), Scottish Gaelic (`gla`).
|
| * **Southeast Asian & Middle Eastern Languages:** Thai (`tha`), Burmese (`mya`), Khmer (`khm`), Lao (`lao`), Tagalog / Filipino (`fil`), Cebuano (`ceb`), Persian / Farsi (`pes`), Hebrew (`heb`), Pashto (`pus`), Kurdish (`kmr`/`ckb`), Uyghur (`uig`), Kazakh (`kaz`), Uzbek (`uzb`), Azerbaijani (`aze`).
|
|
|
| * **Americas & Indigenous Languages:** Quechua (`que`), Guarani (`grn`), Aymara (`aym`), Nahuatl (`nah`), Navajo (`nav`), Mayan languages (`myn`), Inuktitut (`iku`), Cherokee (`chr`).
|
|
|
| > [!TIP]
|
| > **Universal Language Prompting:**
|
| > To translate into any supported language, prepend `<2{iso_code}>` to your source text (e.g., `<2es>` for Spanish, `<2hi>` for Hindi, `<2fr>` for French, `<2de>` for German, `<2ml>` for Malayalam). All 500+ ISO-639 language codes are mapped directly into the model vocabulary and registered in the metadata above for automatic Hugging Face search filtering.
|
|
|
| ---
|
|
|
| <a id="mythos20-model-collections"></a>
|
| ## Mythos2.0 Model Collections
|
|
|
| Following the modular design of frontier foundation families like Qwen, the **Mythos2.0 Series** spans foundation models, specialized context engines, and quantized edge runtimes:
|
|
|
| | Collection / Model | Architecture | Parameters (Total / Active) | Context Window | Target Capability | Status |
|
| | :--- | :--- | :---: | :---: | :--- | :---: |
|
| | **[Mythos2.0-2B](https://huggingface.co/AdithyanAI/Mythos2.0-2B)** | Sparse MoE (8E, Top-2) | **2.04B / 678.7M** | **8,192** | Flagship Universal 500+ Language Translation | 🟢 **Active Run** |
|
| | **Mythos2.0-4B** | Sparse MoE (16E, Top-2) | **4.10B / 1.10B** | **16,384** | Long Document and Legal/Technical Translation | 🔄 *Pipeline* |
|
| | **Mythos2.0-Edge-2B** | 2-bit / 4-bit SMoE | **2.04B (~1.2 GB RAM)** | **4,096** | Sub-2-bit Edge and Mobile Phone Deployment | 📋 *In Dev* |
|
| | **[Mythos-Tokenizer](https://huggingface.co/AdithyanAI/Mythos2.0-2B/blob/main/multilingual_tokenizer.json)** | Byte-Level BPE | **128,000 Vocab** | - | Balanced Compression for 552 Global Languages | 🟢 **Available** |
|
| | **Mythos-16B-Corpus** | Parallel Bilingual Corpus | **16 Billion Pairs** | - | Bicleaner & LASER Curated Parallel Training Data | 🟢 **Open Data** |
|
|
|
| Explore all models in the official Hugging Face Collection:
|
| 👉 **[https://huggingface.co/collections/AdithyanAI](https://huggingface.co/collections/AdithyanAI)**
|
|
|
| ---
|
|
|
| ## Key Features of Mythos2.0-2B
|
|
|
| 1. **500+ Global Languages Supported:**
|
| Native, high-fidelity translation across major world languages plus 250+ underserved African, Indigenous American, and Regional South/Central Asian languages with zero coverage in commercial translation APIs.
|
| 2. **Sparse Mixture-of-Experts Efficiency:**
|
| Employs 8 SwiGLU experts with Top-2 routing. With **2.04B total parameters**, only **678.7M parameters are activated per token**, delivering the translation capacity of a 7B-class model with the inference speed and memory footprint of a sub-1B model.
|
| 3. **8k Native Context with Document Packing:**
|
| Features a native **8,192-token context window** (4,096 encoder + 4,096 decoder) with Block-Diagonal Attention Packing. Translates whole articles, SRT/VTT subtitles, and markdown documents without chunking or losing discourse context.
|
| 4. **FP32 Master Precision Embeddings:**
|
| Maintains a 131M-parameter shared 3-way tied embedding table (`src_embed`, `tgt_embed`, `proj.weight`) in full 32-bit FP32 master weights, ensuring stable representation across rare scripts.
|
|
|
| ---
|
|
|
| ## Comparison with Frontier Models
|
|
|
| | Specification | **Mythos2.0-2B** | TranslateGemma-7B | NLLB-200 (3.3B) | Google Cloud API |
|
| | :--- | :---: | :---: | :---: | :---: |
|
| | **Architecture** | **Sparse MoE (8E, Top-2)** | Dense Transformer | Dense Enc-Dec | Proprietary LLM |
|
| | **Total Parameters** | **2.04B** | 7.0B | 3.3B | Closed |
|
| | **Active Parameters / Token** | **678.7M** | 7.0B | 3.3B | Closed |
|
| | **Context Window** | **8,192 tokens** | 2,048 tokens | 1,024 tokens | Dynamic |
|
| | **Supported Languages** | **500+** | 55 | 200 | 189 |
|
| | **Min Inference VRAM** | **~4 GB** | 16 GB | 8 GB | Cloud API |
|
| | **License** | **Apache 2.0 (100% Free)** | Community License | CC-BY-NC 4.0 | Paid Metered API |
|
|
|
| ---
|
|
|
| <a id="model-architecture-overview"></a>
|
| ## Model Architecture Overview
|
|
|
| * **Model Family:** Mythos2.0
|
| * **Model ID:** `AdithyanAI/Mythos2.0-2B`
|
| * **Architecture:** Encoder-Decoder Sparse Mixture-of-Experts (SMoE)
|
| * **Total Parameters:** 2.04B (`2,037,643,264`)
|
| * **Active Parameters / Token:** 678.7M (`678,688,768`)
|
| * **Layers:** 24 Transformer Blocks (12 Encoder Layers + 12 Decoder Layers)
|
| * **Hidden Dimension ($d_{\text{model}}$):** 1,024
|
| * **Attention Mechanism:** Grouped Query Attention (GQA)
|
| * 16 Query Heads, 4 Key-Value Head Groups (4x KV compression)
|
| * Head Dimension: 64
|
| * RoPE Base Frequency: $\theta = 100,000.0$
|
| * **Feed-Forward Network (Sparse MoE):**
|
| * 8 SwiGLU Experts per layer (Intermediate Dim: 3,072)
|
| * Top-2 Routing with Switch-Transformer Capacity Factor (1.35)
|
| * Calibrated Load Balancing Loss (`aux_loss_weight = 0.01`)
|
| * **Context Capacity:** 8,192 tokens (4k Source + 4k Target Document-Packed)
|
| * **Vocabulary:** 128,000 Byte-Level BPE Tokens (552 languages)
|
|
|
| ---
|
|
|
| <a id="quickstart-free-offline-inference"></a>
|
| ## Quickstart: Free Offline Inference
|
|
|
| ### 1. Installation
|
| ```bash
|
| pip install torch transformers tokenizers sacrebleu
|
| ```
|
|
|
| ### 2. Python Inference Code
|
| ```python
|
| import torch
|
| from tokenizers import Tokenizer
|
|
|
| # Load Tokenizer
|
| tokenizer = Tokenizer.from_file("multilingual_tokenizer.json")
|
| sos_id = tokenizer.token_to_id("[SOS]")
|
| eos_id = tokenizer.token_to_id("[EOS]")
|
| pad_id = tokenizer.token_to_id("[PAD]")
|
|
|
| def translate(model, text: str, tgt_lang: str = "fra", max_len: int = 128, device: str = "cuda:0"):
|
| clean_tgt = tgt_lang.split("_")[0]
|
| prompt = f"<2{clean_tgt}> {text}"
|
| tokens = tokenizer.encode(prompt).ids
|
| src_tensor = torch.tensor([tokens], dtype=torch.long, device=device)
|
| src_mask = (src_tensor != pad_id).unsqueeze(1).unsqueeze(2)
|
|
|
| with torch.no_grad():
|
| enc_out = model.encode(src_tensor, src_mask)
|
| gen_tokens = torch.tensor([[sos_id, tokenizer.token_to_id(f"<2{clean_tgt}>")]], device=device)
|
|
|
| for _ in range(max_len):
|
| cur_len = gen_tokens.size(1)
|
| causal_mask = torch.tril(torch.ones((cur_len, cur_len), dtype=torch.bool, device=device)).unsqueeze(0).unsqueeze(0)
|
| dec_out = model.decode(gen_tokens, enc_out, src_mask, causal_mask)
|
| logits = model.project(dec_out[:, -1:])
|
| next_token = logits.argmax(dim=-1).item()
|
| if next_token == eos_id:
|
| break
|
| gen_tokens = torch.cat([gen_tokens, torch.tensor([[next_token]], device=device)], dim=1)
|
|
|
| return tokenizer.decode(gen_tokens[0].tolist()[2:])
|
| ```
|
|
|
| ---
|
|
|
| ## Training Infrastructure and Engineering
|
|
|
| * **Distributed Engine:** PyTorch Fully Sharded Data Parallel (FSDP) and DDP.
|
| * **Precision Policy:** Native 32-bit FP32 Master Weights with FP16 compute and unscaled FP32 logits projection.
|
| * **Zero-Host RAM Footprint:** Streaming disk-spooler architecture keeping host CPU memory strictly below < 1.0 GB throughout training.
|
| * **Router Stabilization:** Switch-Transformer dynamic capacity factor capping (`capacity_factor = 1.35`) with calibrated auxiliary loss (0.01) preventing expert collapse.
|
|
|
| ---
|
|
|
| <a id="community-and-collaboration"></a>
|
| ## Community and Collaboration
|
|
|
| We welcome researchers, linguists, and compute sponsors to join the Mythos AI initiative:
|
|
|
| * **Compute Sponsors:** Pooling idle GPU hours (RTX 3090/4090, A100, H100) to scale Mythos2.0-4B and Mythos2.0-7B pre-training.
|
| * **Researchers and Engineers:** Optimizing routing loss, sparse kernels, and sub-2-bit quantization.
|
| * **Native Linguists:** Auditing translation quality and expanding low-resource parallel corpora.
|
|
|
| ---
|
|
|
| ## Citation
|
|
|
| If you use Mythos2.0-2B in your research or applications, please cite:
|
|
|
| ```bibtex
|
| @misc{mythos2026multilingual,
|
| author = {Adithyan AI and Community Contributors},
|
| title = {Mythos2.0-2B: A Free and Open-Source Sparse Mixture-of-Experts Translation Foundation Model for 500+ Languages},
|
| year = {2026},
|
| publisher = {Hugging Face},
|
| howpublished = {\url{https://huggingface.co/AdithyanAI/Mythos2.0-2B}}
|
| }
|
| ```
|
|
|
